, , _ mm‘un' ' ' ' " ' ‘ ... " " ‘~ ' ‘w' . ’ ' ., .-*-~..-v:.-- ‘- --.. r... ' ‘ “"4 v ' ’ .' 4%. “ ' r?» W:- , . . z. .. 'h....{..1._.. . .. ‘ .. ., ., ‘ , , v ”L - .....n, ' . ,0. > - . . . , . K , , fl , , .~\. ' .. . ‘ ' '~ A“ -~t.e..‘- 4; V ._ 2. f r. “5"?“0’! . I. . , , ., H, ,. 44. . ,. {5“ 'k 4 . . 'rl“ ML}: 1’.: .1.” ~.. ly— .‘LA «I _ 1+; tar}. ,;c L. 30;- -t» “.‘Y‘..‘.‘r"‘,:..3‘: “r: "5" ‘5“ he“ . .31 fl . L‘, (”1 Ctt‘ w. 5 . a, . , . . ; :33, 7:47“? .‘l— : . #316. . . . ‘7‘?” -1 .4 .d R < I‘- ‘4 . . . m1. .. . . ’2 d ._, L , ....-..;:."‘ .. . .. .4. 14-4un-F ”2:55"; .wptis' '. — k “at‘ uawxs- wiwmrg. 3.92:3;1 ‘9 “a" -. '14 .66. 5.10553“: 1 . . ,. J; ‘1' :1“... " 1x ' ~‘ “" P. M n s ;. ~ ~ m... 4. H, ‘- .“P-z .: «nu .15! " "*w I“ ~ ' L ' ‘ m “fig“‘fiv‘y ”#51. iii?" a. mum?“ m2: 331;? ‘ M *‘- . J a. . . 2:,”- , .82.,é: J.“ “A. , ”:9.“ " 5:“ "' A. h ‘ s' I . ‘ I ., . ‘35.??? 3% V r . A . . V r’r‘iififiv" .‘fig . _ - , . 92.2%., « . " if ' ”I -".' 4 a”; , . us mlg-J 4 . . - . ‘ ~<‘-" ' 4 H“ . ‘ .4: .4 a K 1:12. NEZW 5‘ I '0: a,“ ‘7,“ ‘ rm: my, fit... ‘ . » ~ - - .1 . ‘11“ hit”..- L 1.5.3; EZREH'LGE, *vK ; {.1, ‘ ,,.. . a.» . . .y ‘W ”ma-f, ‘54: L’" M ‘1 fir“ .. .. , g .' . .. . . ‘3‘ nm .5 4 - ,n V _ 3. “A; u , _ -'L-‘.’ 42-3-0 1 — «m 33.4%.: ~ A5; it,§..un-.-v ”5;: . . -r~r~ .4 H {. J, I“ V‘ . .2. . V 3. . 4., . .- ._ 35:“ - u: f: fl. . .. 'qrmkfi mr . ,. - , ”fifty-3’3 :- ~ ~ I. p . . d _ . \: . _ . . , ‘ .. .., . . L. 1‘ ' : -. » , ‘ ""Hh'h 1" ’ u '. ‘1 ‘fll ' ' flirt}: :32. ' ' r‘§r$7."g§§¢.mz:u {'1‘ ‘ 1 x1 > . (1.“: 4w. ~ “v- - ~ “*3“ M ' u.~.r “4G". w, -r - , "f ‘ - 75-} :t’;..::.:f ~ ' U « 1.1.? h 7"- ..,g.L‘3,“':.- m . ,4 :: '.. A, .g. ' ‘er ",1“. : s" , '7. ,» , "£22.54; ' .. p. . . u }.' A ’ r ‘ I :’ t 243‘ .3 . I, - v, ’3' by." ‘j 1 3] € r i r a C . f‘T-V'h-u ' ENC ml . . 1" a ‘v ’ i ‘33. ' - w- . 4‘” ' ’ .‘11 fl ‘, )1 V. I; , l 1" 7 '3 k". ‘N‘l ,. '7 .9- i. ’. u“ x h, 1‘ a J”. ”l" 'ld‘fl; ’3 1., . ‘ IJ}\?I‘rKv‘h%’,w Iii"); finrv. ‘4: r3,” _ 1;;an a}??? $1")? MICHI IGAN STATEU II I III IIIIIIIIIIIIIIIIIIIIIIIIIIII 00877 2372 II This is to certify that the thesis entitled Effect of directly acidified Cottage cheese whey ultrafiltration retentates on the physical and sensory properties of orange sherbet. presented by FLORA GEORGIOS MANGANARI has been accepted towards fulfillment of the requirements for M.S. . Food Science degree in Qgflaéw Major professor Date 2'27‘ ?Z 0-7639 MS U i: an Affirmative Action/Equal Opportunity Institution ’———— w ~77 ,7 ,7 ———- - _,_ I LIBRARY 4 I Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE MSU Is An Affirmetlve Action/Equal Opportunity Institution cMma-nt EFFECT OF DIRECTLY ACHJIFIED CO’l'l‘AGE CHEESE WHEY ULTRAFILTRATION RETENTATES ON THE PHYSICAL AND SENSORY PROPERTIES OF ORANGE SHERBET by FLORA GEORGIOS MAN GANARI A THESIS M10111 SubsmitteilI to ' ' an tate niversity in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science and Human Nutrition 1992 r“) fix. 673-323“ ABSTRACT EFFECT OF DIRECTLY ACIDIFIED COTTAGE CHEESE WHEY unnmununuummvnmnnnmmmsmNTHEPHmmCALAND SENSORY PROPERTIES OF ORANGE SHERBET by FLORA GEORGIOS MAN GANARI Directly acidified Cottage cheese whey concentrated by ultrafiltration and diafiltration was used as an ingredient in orange sherbet. Four sherbet mixes containing 0% (sherbet A), 25% (sherbet B), 50% (sherbet C) and 75% (sherbet D) whey solids as a replacement of milk solids non-fat were prepared. All sherbets otherwise contained 1.5% milkfat, 3.4% solids non-fat, 20% sucrose, 9% corn syrup solids and 0.3% emulsifier-stabilizer. Sherbet A had lower (p<0.05) melting resistance than sherbets containing whey solids. In a consumer panel, sherbet B received a score of 7.02 and 7.42 for the flavor and texture acceptability, respectively, on a 9-point hedonic scale, and was found similar to control sherbet A, but more acceptable in flavor (p<0.05) and texture (p<0.001) than sherbets C and D. Sherbets A and B were found less icy and creamier than sherbets C and D (p<0.01). Sherbet A was found sweeter than sherbets C and D (p<0.05), whereas sherbet B was sweeter than sherbet D (p<0.05). Overall, sherbet B was found not different from control. I dedicate this work to my parents for their love and support. ii ACKNOWLEDGEMENTS The author would like to thank her major advisor, Dr. J. Partridge, for his support and guidance during the time spent at Michigan State University. Special gratitude is also expressed to Dr. D. Ott for her very useful suggestions and help in the sensory evaluation part of this research. Sincere thanks to the State Scholarship Foundation of Greece, which supported me financially. Without their help I would not be able to pursue graduate studies outside Greece. Appreciation is also extended to Dr. B. Haines, Dr. B. Harte and Dr. J. Stefi‘e for serving as members of the guidance committee and for their very helpful comments and suggestions for the improvement of this thesis. Sincere thanks to the faculty and students who participated in the sensory panel training sessions and especially to Sandy Daubenmire, Mustafa Farouk, Enayat Gomaa, Kay Kresl, Chit Leong, Hsiao-Yuan Li, Frank Mona- han, Virginia Vega and Mike Watt who served as members of the trained sen- sory evaluation panels. I would like to thank Mr. Mike Watt also for his help and support throughout my research. A special thanks is extended to Ms R. Cracks], for helping me find enough panelists for the sensory evaluation tests. I would also like to thank Country Fresh Inc, for supplying the acid whey and for sending me any information I requested. Finally, I thank my friends D. Argyropoulo and A. Drakopoulo, who helped me without hesitation whenever I needed any help. LLADL TABLE OF CONTENTS PAGE LIST OF TABLES ............................................................................ vii LIST OF FIGURES .......................................................................... xi ABBREVIATIONS ............................................................................ xii INTRODUCTION ............................................................................. 1 LITERATURE REVIEW .................................................................. 3 DEFINITION AND COMPOSITION OF WHEY ...................... 3 USES OF WHEY ........................................................................ 5 THE NECESSITY OF WHEY UTILIZATION .......................... 8 THE BENEFITS OF WHEY UTILIZATION ............................ 9 MODIFIED WHEY PRODUCTS ............................................... 10 METHODS OF WHEY PROCESSING ..................................... 11 Ultrafiltration .................................................................. 13 Functional properties of whey protein concentrates ...... 17 _ USES OF WHEY IN FROZEN DESSERTS .............................. 19 EXPERIMENTAL PROCEDURE ................................................... 28 PREPARATION OF WHEY PROTEIN CONCENTRATES ...... 28 TOTAL SOLIDS DETERMINATION ........................................ 29 FAT DETERMINATION ............................................................ 29 TOTAL NITROGEN DETERMINATION ................................. 30 NON-PROTEIN NITROGEN DETERMINATION ................... 30 pH DETERMINATION .............................................................. 31 ’ TITRATABLE ACIDITY ............................................................ 31 PREPARATION OF SHERBET MIXES ................................... 31 FREEZING OF THE SHERBET MDCES .................................. 33 OVERRUN DETERMINATION ................................................ 34 DETERMINATION OF THE RHEOLOGICAL PROPERTIES < OF THE SHIERBET MIXES ....................................................... 35 MELTING RESISTANCE TESTS ............................................. 36 SENSORY EVALUATION ......................................................... 36 STORAGE STABILITY .............................................................. 39 HEAT-SHOCK STABILITY ....................................................... 39 PANEL TRAINING .................................................................... 40 Training the iciness panel ............................................... 42 iv CAAA. Training the creaminess panel ........................................ Training the sweetness panel .......................................... RESULTS & DISCUSSION ............................................................. TOTAL SOLIDS & FAT DETERMINATION ........................... TOTAL, NON-PROTEIN & PROTEIN NITROGEN ................ pH AND TITRATABLE ACIDITY ............................................. DETERMINATION OF THE RHEOLOGICAL PROPERTIES OF THE SHERBET MIXES ....................................................... MELTIN G RESISTANCE OF SHERBETS .............................. SENSORY EVALUATION OF SHERBETS ............................. Flavor acceptability of sherbets ...................................... Texture acceptability of sherbets .................................... Iciness, creaminess and sweetness comparison of sherbets ............................................................................ HEAT SHOCK STABILITY TESTS .......................................... STORAGE STABILITY TESTS ......................................... ' ........ COMPARISONS .......................................................................... FURTHER DISCUSSION ON THE PANELISTS’ . COMMENTS ................................................................................ CONCLUSIONS AND RECOMMENDATIONS ............................ APPENDIX A ................................................................................... DRY WHEY ANALYTICAL STUDY ......................................... APPENDIX B ................................................................................... PRODUCTION AND UTILIZATION OF WHEY ..................... APPENDIX C ................................................................................... QUESTIONNAIRES FOR THE SENSORY EVALUATION TESTS ......................................................................................... APPENDIX D ................................................................................... HYPOTHESES FOR THE SENSORY EVALUATION TESTS ......................................................................................... APPENDIX E ................................................................................... WORKSHEETS FOR THE SENSORY EVALUATION APPENDIX F .................................................................................... VERBATIM FROM THE SENSORY EVALUATION TESTS... 'Verbatim for the flavor of sherbet A (control) .............. 'Verbatim for the texture of sherbet A (control) ............ 'Verbatim for the flavor of sherbet B ............................. 'Verbatim for the texture of sherbet B .......................... 'Verbatim for the flavor of sherbet C ............................. 'Verbatim for the texture of sherbet C .......................... 'Verbatim for the flavor of sherbet D ............................. 'Verbatim for the texture of sherbet D .......................... APPENDDC G ................................................................................... V PAGE 45 47 49 49 50 ' 51 52 58 61 61 63 71 72 74 78 80 83 83 86 86 89 89 93 93 95 95 104 104 104 106 107 109 110 112 114 116 118 PAGE RHEOLOGICAL RAW DATA (SHEAR STRESS / SHEAR RATE FOR THE SHERBET MDCES ......................................... 118 APPENDIX H ................................................................................... 120 AN OVA TABLES FOR OBJECTIVE AND SENSORY TESTS ......................................................................................... 120 TABLE 10 11 12 13 LIST OF TABLES Composition (%) of whey produced by difl'erent methods of casein precipitation ........................................ Characteristics of food processing wastes ...................... Composition changes of whey occurring during ultra- filtration ............................................................................ UF membrane materials and their properties ............... Formulation of the sherbet mixes ................................... Ingredients used in the formulation of sherbet mixes Amount of ingredients used for 36.29 kg of sherbet mix A, B, C and D ............................................................ Composition and treatments of sherbets used for the iciness panel training ....................................................... Tests performed for the iciness panel training .............. Composition of the samples used for the creaminess panel training ................................................................... Tests performed for the creaminess panel training ....... Composition of the samples used for the sweetness panel training ................................................................... Tests performed for the sweetness panel training ......... PAGE 15 16 ' 32 33 33 43 46 46 47 TABLE 14 15. 16 17 18 19 20 21 22 23 24 25 26 27 28 Total solids and fat content of acid whey, WPC, sherbet mixes and sherbets ............................................. Percentage of total (TN), non-protein (NPN) and protein nitrogen (PN) of whey, WPC, sherbets and sherbet mixes ................................................................... pH and titratable acidity of acid whey, WPC, sherbet mixes and sherbets .......................................................... Rheological constants for the linear and the power law model at 40°F (4.44°C) .............................................. Tukey’s test results for the viscosities of sherbet mixes at 40°F (4.44°C) ..................................................... Tukey‘s test results for the melting resistance of sherbets at 38°C (100.4°C) ............................................. Tukey’s test results for the flavor acceptability of sherbets after 8 days of storage ....................................... Percentage number of responses in each category of the 9-point hedonic scale for the flavor of each sherbet ........ Tukey’s test results for the texture acceptability of sherbets after 8 days of storage ....................................... Percentage number of responses in each category of the 9-point hedonic scale for the texture of each sherbet ..... Tukey’s test results for the iciness of sherbets after 9 days of storage .................................................................. Tukey’s test results for the sweetness of sherbets ......... Tukey’s test results for the creaminess of sherbets ....... Tukey’s test results for the flavor and texture acceptance and the iciness of sherbets after a heat shock treatment ............................................................... Tukey’s test results for the flavor acceptance, texture acceptance and iciness of sherbets after storage ............ 0.. PAGE 49 51 52 55 56 58 62 63 66 67 69 70 72 73 TABLE 29 A1 A.2 B.1 B.2 C.1 C.2 C.3 E.1 E.2 E.3 E.4 E.5 E.6 E.7 G.1 Tukey‘s test results for the iciness of sherbets after 9 and 32 days of storage and after a heat shock treatment .......................................................................... Average vitamin, mineral and amino acid contents of sweet-type dry whey ........................................................ Average vitamin, mineral and amino acid contents of acid-type dry whey ........................................................... Estimated U.S. fluid whey and whey solids production by type and resulting quantity of whey solids further processed ........................................................................... Utilization of whey and whey products in animal feeds. Comparison of 1989 and 1988 end-uses .......................... Questionnaire (1) for the flavor acceptance test ............ Questionnaire (2) for the flavor acceptance test ............ Questionnaire for the iciness comparison test ............... Flavor acceptance test worksheet ................................... Texture acceptance test worksheet ................................. Iciness comparison test worksheet (test performed after 9 days of storage) .................................................... Sweetness comparison test worksheet (test performed alter 10 days of storage) ................................................. Creaminess comparison test worksheet (test performed after 11 days of storage) ............................... Iciness comparison test worksheet (test performed after a heat shock treatment) .......................................... Iciness comparison test worksheet (test performed alter 32 days of storage) .................................................. Shear rate and shear stress data for sherbet mixes at 40°F (4.44°C) ...... PAGE 78 84 85 87 88 90 91 92 96 97 98 100 101 102 103 119 TABLE H.1 H.2 H.3 H.4 H.5 H.6 H.7 H.8 H.9 H.10 H.11 H.12 H.13 H.14 H.15 AN OVA table for the viscosities of sherbet mixes .......... AN OVA table for the melting resistance of sherbets at 38°C (100.4°C) .................. AN OVA table for the flavor acceptance of sherbets after 8 days of storage ...................................................... AN OVA table for the texture acceptance of sherbets after 8 days of storage ...................................................... AN OVA with interaction table for the iciness test after 9 days of storage ............................................................... AN OVA table for the Schefl'e paired comparison test for the iciness of sherbets after 9 days of storage ................ AN OVA table for the Schefi'e paired comparison test for the sweetness of sherbets ............................................... AN OVA table for the Scheffe paired comparison test for the creaminess of sherbets .............................................. AN OVA table for the flavor acceptance of sherbets alter a heat shock treatment .................................................... AN OVA table for the texture acceptance of sherbets after a heat shock treatment ............................................ AN OVA table for the Schefi'e paired comparison test for the iciness of sherbets after a heat shock treatment ..... AN OVA table for the flavor acceptance of sherbets afier 31 days of storage ............................................................ AN OVA table for the flavor acceptance of sherbets after 122 days of storage .......................................................... AN OVA table for the texture acceptance of sherbets after 31 days of storage ................................................... AN OVA table for the Schefl'e paired comparison test for the iciness of sherbets after 32 days of storage .............. X PAGE 120 120 121 121 121 122 122 123 123 123 124 124 124 125 125 LIST OF FIGURES FIGURE PAGE 1 Schematic representation of the ultrafiltration of whey ................................................................................. 14 2 Torque versus time diagram for sherbet mix C ............ 53 3 Melting resistance of sherbets at 38°C (100.4°F).. ........ 59 4 Flavor acceptance of sherbets after 8 ( ), 31 (El ), 122 ( IIIII) days of storage and after the heat shock treatment ( ). (1 = dislike extremely, 9 = like extremely on the 9-point hedonic scale) .......................... 75 5 Texture acceptance of sherbets after 8 ( ) and 31 ( ) days of storage and after the heat shock treatment ( ). (1 = dislike extremely, 9 = like extremely on the 9-point hedonic scale) .......................... 76 MOAT—LDC .......... I. BCCCCCDDEGGEm thNthPFFEEIIIII AN OVA: BOD: CAS: CF: CMC: COD: CSS: DE: DSW: ED: GF: GLC: MS: MSNF: MSU: NFDM: NPN: PER: PN: RO: RPM: STEM: TA: TCA: ABBREVIATIONS Analysis of Variance Biological Oxygen Demand Sodium Caseinate Concentration Factor Carboxy-Methyl-Cellulose Chemical Oxygen Demand Corn Syrup Solids Dextrose Equivalent Dry Sweet Whey Electrodialysis Gel Filtration Gas Liquid Chromatography Ion Exchange Maximum Minimum Mass Spectroscopy Milk Solids Non Fat Michigan State university Non Fat Dry Milk Non Protein Nitrogen Protein Efficiency Ratio Protein Nitrogen Reverse Osmosis Revolutions Per Minute Solids Non Fat Stabilizer/Emulsifier system Titratable Acidity Trichloroacetic Acid Total Nitrogen mtrafiltration Whey Protein Concentrate :26 ie rn Chapter 1 INTRODUCTION Several studies that have been done on the utilization of whey indicat- ed that whey is a good source of solids for frozen desserts. In most of the stud- ies, however, dry whey or whey protein concentrates from sweet whey were ex- amined. Few researchers have worked with acid whey, so the information for the utilization of this by-product as a frozen dessert ingredient is limited. In this study, acid whey from direct-set Cottage cheese was used to supply part of the solids in an orange sherbet formulation. Casein precipita- tion in the cheese was achieved by acidifying with hydrochloric acid (HCl). Di- rect-set Cottage cheese was chosen, because it does not contain any added cul- tures that could impart fermented flavors to the whey and, possibly, to the final product. There appeared to be a greater potential for use of direct-set acid whey in foods than for culture-set acid whey. Sherbet was chosen as the frozen dessert for this research for two reasons. The acidity of such a whey would be compatible with the fruit flavor sherbets have. Also, about forty six million gal of sherbet are produced every year in U.S.A. [IAICM (1984)] and a 25% re- placement of the milk solids non-fat of these sherbets by acid whey solids would consume about 60 million lbs of acid whey (from an average of 5,700 mil- lion lbs produced each year - Table B.1.). The sherbet was chosen to have or- ange flavor, because orange sherbet is traditionally the most popular sherbet in U.S.A. [IAICM (1965)]. These tose zau'on tion ar placer! acid is" lance, finalp 2 The acid whey was fractionated by ultrafiltration and diafiltration. These processes concentrated the whey protein while removing most of the lac- tose and minerals from the acid whey. The objective of this study was to determine the feasibility of the utili- zation of acid whey from direct-set Cottage cheese concentrated by ultrafiltra- tion and diafiltration as an ingredient in orange sherbet. The effect of the re- placement of milk solids non-fat at the level of 25, 50 and 75% with fractionated acid whey on the rheological properties of the sherbet mixes, the melting resis- tance, the organoleptic quality and the storage and heat-shock stability of the final products were determined. DEF caseir chees whey Swee‘ 0f re! whey norm acid add. Whey (197g chees lion ( Table [ADP and Chapter 2 LITERATURE REVIEW DEFINITION AND COMPOSITION OF WHEY Whey is the greenish-yellow liquid remaining after the precipitation of casein and removal of fat from milk. It is a by-product in the manufacture of cheese and may be classified as sweet whey, known also as rennet whey, or acid whey, depending on the kind of coagulation used in the cheesemaking process. Sweet whey is from the manufacture of cheese or casein from milk by the action of rennet-type enzymes with relatively little or no acid development. Acid whey is produced when milk is coagulated primarily by acid. Acids that are normally used for the cheese coagulation are food grade lactic acid, sulfuric acid, hydrochloric acid, phosphoric acid, D-glucono-delta-lactone and citric acid. Manufacture of Cottage cheese results in acid whey production. Sweet whey has a minimum pH of 5.6 and acid whey a maximum pH of 5.1 [Arbuckle (1979), Hansen (1979), Marshall (1982), Sienkiewicz and Riedel (1990)]. The composition of whey varies with the composition of milk, the cheese or casein type and the processing methods. The approximate composi- tion of whey produced by difi'erent methods of casein precipitation is given in Table 1. In addition to the constituents mentioned in Table 1, another study [ADPI (1991)] includes mineral, vitamin and amino acid composition of sweet and acid dry wheys. The results of this analysis are presented in Appendix A (Table I Hedricl Table I pH of] Total I Lactos Fat Total j Caseii Whey Ash Low-r N ton LaCtlt l and h eentr 0958i) (1982 4 (Table A.1, A.2). Comparison of these results with those found by Glass and Hedrick (1977a,b) shows very good agreement. Table 1. Composition (%) of whey produced by different methods of casein precipitation [Hansen and Jensen (1977)]. Rennet precipitation Acid Precipitation Lactic acid Biological Chemical fermentation Lactic acid HCl H2804 Lactic bacteria acid pH of precipitation 6.4 5.4 4.6 4.5 4.5 4.5 Total solids 6.27 6.43 6.00 6.40 6.44 6.70 Lactose 4.79 4.56 3.93 4.81 4.78 4.80 Fat 0.04 0.04 0.04 0.04 0.04 0.04 Total protein 0.82 0.87 0.80 0.70 0.70 0.70 Casein 0.19 0.15 0.17 0.16 0.17 0.16 Whey proteins 0.41 0.46 0.35 0.36 0.35 0.37 Ash 0.48 0.63 0.65 0.78 0.79 0.68 Low-molecular wt. N-compounds 0.22 0.26 0.28 0.18 0.18 0.18 Lactic acid 0.14 0.33 0.62 0.13 0.13 0.53 The major protein constituents of whey are B—lactoglobulin, a—lactalbu- _ min, bovine serum albumin, immunoglobulins and proteose-peptones. There are several minor whey proteins including lactoferrin, lactollin, glycoprotein and blood transferrin. Whey fi-om bovine milk contains 4-7g protein/L, the con- centration depending on the type of whey, the stage of lactation and the pro- cessing conditions used in the manufacture of cheese or casein [Marshall (1982)]. USES produc its peri intoda 1988, I whey l produi mm m. pleme P0115 cattle kiewi tion c bush. am] powd been mum ids ,1 [Mat] USES OF WHEY Manufacture of cheese, Cottage cheese or industrial casein results in production of up to 9 kg of liquid whey for every kg of final product. Because of its perishable nature, whey cannot be easily stored for any length of time, and in today’s environment its potential as a major pollutant prohibits dumping. In 1988, 55,776 million lbs of fluid sweet and acid whey and 3,625 million lbs of whey solids were produced in US. Estimated U.S. fluid whey and whey solids production (by type) and resulting quantity of whey solids further processed are summarized in Table B1 in Appendix B [ADPI (1991)]. Whey and whey products are being used in animal feeds, as fertilizer on the land and in human foods. One of the first uses of fluid whey was to sup- plement the vitamin and mineral diets of poultry and pigs. Some research re- ports also indicated that liquid whey was an acceptable feed for dairy and beef cattle. Today, whey is used as part of the feed of many different animals [Sien- kiewicz and Riedel (1990)]. Table B.2 in Appendix B summarizes the utiliza- tion of whey and products derived from it in animal feeds for 1988 and 1989 [ADPI (1991)]. As a fertilizer, whey was found to increase corn yields by 110 bushels to the acre, when it was sprayed at a rate of 8 in/acre (one inch to the acre represents 22,000 gallons) [Ryder (1980)]. An enhancement in “mouthfeel” has also been reported, when whey powder or dried whey are used as ingredients in foods. Whey products . have been used to replace part of the non-fat dry milk normally used in preparing commercial foods [Seal (197 6)]. In infant food formulations, modified whey sol- ids may be added to bovine milk to give it characteristics of human milk [Mathur and Shahani (1979)]. Liquid, concentrated or spray dried whey are also used in frozen desserts. Whey used for frozen desserts should be prefera- bhrni theci hhhu aMe itwas udcc ass] addi whey U 77 madc iunt [Ann ness. finds fleas (NFD bette: in th. (1979 milled thl 6 bly made from single culture W. Organisms which convert the citrates to diacetyl or other flavor compounds are undesirable in ice creams [Arbuckle (1979)] . Limited information concerning liquid whey use is avail- able. Its use as an ingredient for flavored beverages has been investigated, and it was found that a 100% substitution of Cottage cheese whey for water as a liq- uid component resulted in equal acceptability. Acid whey has also been used as a base for salad dressings and was well accepted by a sensory panel. The acid whey enhanced the tartness of the salad dressings. In acid products, acid whey imparts desirable flavor characteristics [Holmes (1979), Stull et al. (1977 )1. Liquid cheese whey has been used directly from the Cottage cheese op- eration in the manufacture of ice cream. This acid whey is usually standard- ized to titratable acidity 0.13-0.14 or to pH 6.6-6.7 before use in ice cream mix [Arbuckle (1979). Hansen (1979)]. Whey may be used as a tenderizer and helps retain moisture and fresh- ness. It helps to produce a better crust for pie dough and softer textured baked goods with longer shelf life in grocery stores and bakeries. Whey gives a more pleasing color to almost any bakery product than when non-fat dry milk (NFDM) is used alone. Bread with whey as an ingredient was said to have been better-colored and smoother-textured. Sweet rolls and coffee cakes with whey in the dough had more even browning and more acceptable flavor [Holmes (1979), Mathur and Shahani (1979)]. Ice creams, fudges, toppings, caramels, syrups, coatings, frozen pies and fillings were said to have been better products as a result of the use of whey. The blending of candies and their appearance are said to be improved when whey is used in their formulation [Saal (197 6)]. Some of the large flour companies experimented with whey in their mixed products and had good results. Kraft entered the bake-mix business with products in which a principal ingredient has been dried whey since 1974. Afiery rolls ar tities o capacit proved coating maker cheese ”0011! the Ni hani( mova] Produ Dentrz their that h WPC; Wee, (1979‘ WhEy Both l 7 After years of research, the company came up with special mixes for breads, hot rolls and various kinds of cakes. Whey solids in combination with small quan- tities of gelatin have been advocated as a new land of flow agent, that has the capacity to hold twice its own weight in oils, fats and flavors. This property proved very useful for the production of non-aqueous products. Whey-based coatings have been suitable for food applications and used by ice cream novelty makers, candy makers and bakers [Mathur and Shahani (1979), Sea] (1976)]. The ultimate result of research sponsored by Kraft was Velvetta cheese, with whey as a major ingredient [Saal (1976)]. Some of the commercial products manufactured utilizing whey as an ingredient include breakfast drinks, Ricotta cheese, chip dips, spreads, sour cream, buttermilk, yogurt and the Norwegian cheese Mysost [Holmes (197 9), J elen (1979), Mathur and Sha- hani (1979)]. Whey protein concentrate (WPC) is the substance obtained by the re- moval of sufficient non-protein constituents from whey, so that the finished dry product contains not less than 25% protein [CFR (1990a)]. Whey protein con- centrates have nutritive values and functional properties which determine their utilization. The good solubility of these concentrates over a wide pH range make them good ingredients for beverages containing up to 3% protein that have been prepared over a pH range of 2.5-7.0. Other products made with . WPC are marshmallows and similar confectionery items, desserts of the souffle type, meringues and frozen desserts [Anon (1979), Marshall (1982), Mathur (1979). Muller (1976)]. Brothiness, bitterness and volatile acidity flavor characteristics of acid whey from direct-set and cultured products may limit use in bland products. Both the volatile and the non-volatile fractions contribute to these flavors. The amino acids, peptides and calcium salts probably contribute to the brothy and litter f Iwhey-I THEI recent] stream for use amour erly d1 5113th tion. I (BOD: Sister fields liquid 0Xl’ge dllced and 0 {00d I to the 0ther 8 bitter flavors, however there may be additional components that contribute a “whey-like” flavor to bland products [McGugan et al. (1979)]. THE NECESSITY OF WHEY UTHJZATION Until the 20th century, nobody cared what happened to whey. Until recently, the methods of disposal involved dumping the whey into a sewer or stream, giving it back to the farmer for feeding to hogs, or spreading it on fields for use as a fertilizer. People have realized that whey, discharged in large amounts into rivers and streams, was a fatal pollutant for fish. When improp- erly dumped in the fields, it gave off a rank, fetid, skunk-er odor. People per- suaded their politicians to enact legislation prohibiting these kinds of pollu- tion. Stringent limits were placed on the volume of biological oxygen demand (BOD) that a particular plant could discharge into municipal or county sewer systems, and whey, had to be properly diluted, before it could be sprayed on fields [Christensen (1976), Saal (1976)]. It is worth mentioning that 100 Kg of liquid whey, containing approximately 3.5 Kg of BOD and 6.8 Kg of chemical oxygen demand (COD), has the polluting strength equivalent to sewage pro- duced by 45 people [J elen (197 9)]. Whey is the most potent of all dairy wastes and one of the strongest wastes of any kind. Table 2 shows the BOD of several food processing wastes, including whey. Dairy plants had to build treatment facilities at great cost. This added to the cost of manufacture per pound of cheese. There were not enough pigs or other animals to economically absorb the available whey supply as slops, so cheesemakers have redoubled their efl‘orts to find new uses for whey. Tabl Dair) (hhe the m Untrie ityto quant, Office teins_ “938 of Careal 3.2, w] 9 Table 2. Characteristics of food processing wastes [J elen (1979)]. W W Dairy processing waste waters fluid milk plant 1000 ice cream plant 2500 Cottage cheese plant 6000 whey powder plant 40 Other food processing waste waters sweet goods bakery ' 2500 meat canning 1500 candy plant 4000 poultry processing 5000 Raw wastes sweet whey 35000 acid whey 45000 fish processing stickwater 50000 domestic sewage 300 THE BENEFITS OF WHEY UTHJZATION Although whey is considered a waste product, it contains about 20% of the milk protein, almost all of the milk sugar, and altogether about 50% of all nutrients consumed normally in milk [J elen (1979)]. Cheese whey has the abil- ity to supply high amounts of whey protein. The quality of any protein is de- termined by the lowest quantity of any one of the essential amino acids. A high quality whey protein consists of at least 18 amino acids. It has an oversupply of five of the seven essential amino acids that are usually lacking in other pro- teins. Whey protein provides a good fortifying effect for most foods as the ex- cess of these five amino acids complement deficiencies, such as those found in cereal grains. The average Protein Eficiency Ratio (PER) of whey protein is 3.2, whereas soy protein has PER 1.8 and casein 2.5. Ten pounds of whey blende 2.23. ' proteix (1977): ings in increa lated l the ar non-fa MOD er pro, lactosr Femov acidity modjfi app“); 10 blended with ninety pounds of soy protein will increase the PER of the blend to 2.23. Whey protein concentrates were found to be very good supplementary proteins for soy protein and flour [Loewenstein (1975), Muller (1976), Weiner (1977)]. , Whey is very cheap, and manufacturers can realize considerable sav- ings in replacement of milk solids by whey solids. The price of milk solids keeps increasing and, consequently, so does the price of the products that are formu- lated with this ingredient. This is. the reason why there is a lot of research in the area of substitution of whey solids in frozen dairy products for milk solids non-fat (MSNF) [Frazeur( 1977), Hekmati and Bradley (1979)]. MODIFIED WHEY PRODUCTS One of the major problems hindering the development of new consum- er products containing whey is its high moisture combined with the high salt, lactose and, for the case of acid whey, high acid content. The need for water removal for most product uses accentuates the saltiness, lactose content and acidity even more. There is, however an efl‘ort towards the manufacture of modified wheys, which have composition similar to the MSNF composition or- appropriate for their incorporation into a certain food item. The term modified whey products is used to designate a group of whey products obtained through processing whey by special techniques. Examples of modified wheys are par- tially delactosed whey, partially demineralized whey, demineralized whey and whey protein concentrates [Frazeur (1977), Weiner (1977 )]. Ingredient suppliers to the ice cream manufacturers have been re— structuring their ingredients, to provide the industry with functional products or ing.‘ been d the sa signifi- tein Qt in the ____———— and Si F: with e. and w] appliet MET 11 or ingredients at lower cost. After extensive testing, a series of products has been designed to replace MSNF in frozen desserts. These ingredients provide the same functional characteristics as MSNF, while at the same time offering significant savings potentials. With some of the replacement products, the pro- tein quality, as measured by the PER, is actually increased. Ingredients found in the market include specially processed WPCs with enhanced dairy flavor and sweetness and reduced undesirable after-tastes, formulations of WPC, milk protein and whey solids that provide a balance of product functionality with excellent economics and spray-dried products consisting of milk proteins and whey solids [Anon (1976), Carter et al.(1982)]. Computer simulation pro- grams are available that make cost predictions on any combination of processes applied to whey [Olson (1979)]. METHODS OF WHEY PROCESSING The traditional processes used for the production of dried whole whey include spray drying, roller drying, concentration to semisolid feed blocks or production of sweetened condensed whey [J elen (1979)]. Use of dried whole whey in human foods is limited due to varying functibnal properties of the in- dividual components. The objection to the whey powder is that the protein lev- el is only about 12% in relation to the high lactose and ash content, and this fact limits the amount that can be used in food formulations. Roller and spray dried whole whey powder products were gritty, insoluble and difficult to incor- porate in food products [Christensen ( 1976)]. Fractionation techniques can be used to remove some of the undesirable components (salts, lactose, acids) and recover the most valuable whey components. The relatively recently developed and cor change widene structe whey flows the ' tein n permit branes site res Princi; Phase Whey i ion ex 111% ac throw direct kilo“; IUJOS. 12 and commercialized separation techniques such as electrodialysis (ED), ion ex- change (IE). reverse osmosis (R0) and ultrafiltration (UF) have substantially widened the range for manufacture of various fractionated, modified or recon- structed whey products [J elen (1979), Morr et al. (1973)]. The electrodialysis process reduces the mineral and nitrate contents of whey without any significant effect on lactose and protein contents. The whey flows through the electrodialysis module with ion-selective membranes under the influence of a small electrical potential. Cation selective membranes con- tain negatively charged, covalent-bonded groups, such as sulfonic acid that permit the passage of cations and exclude anions. On the anion selective mem- branes the positively charged groups, quarternary amines, produce the oppo- site result [Christensen (1976), Sienkiewicz and Riedel (1990), Smith (1976)]. Ion—exchange also is used to remove minerals from whey. The basic principle of this technique is the exchange of “mobile” ions of the stationary phase for the equivalently charged ions fi-om the surrounding solution. The whey is first conducted through a cation and then an anion exchanger. The cat- ion exchanger binds the cation of the minerals with the release of correspond- ing acids, whose anions are bound to the anion exchanger. After its passage through both exchanger columns the whey is demineralized, depending on its type, from 90 to 99%. Whey which has been demineralized up to 90% can be directly concentrated and dried [Sienkiewicz and Riedel (1990 )1. Reverse osmosis (R0) is a membrane separation technique (also known as hyperfiltration), in which only water and small amounts of solutes from whey are removed, resulting in concentration of the total solids. R0 in- volves the use of semi-permeable membranes (polyamide, polysulfone and cel- lulose triacetate membranes) with pore size 0.4 nm and hydraulic pressure usuall and Hi tose a1 memb‘ 1,000 I Ultrz 13 usually between 50-100 bar (5,000-10,000 kPa) [Marshall (1982), Sienkiewicz and Riedel ( 1990)]. In ultrafiltration (UF), a membrane permeable to water and small mol- ecules, but not to large molecules, separates the protein from the smaller lac- tose and salt molecules. Ultrafiltration involves the use of semi-permeable membrane with pore size of about 2 nm and pressure between 1-10 bar (100- 1,000 kPa) [Anon ( 1979), Sienkiewiczand Riedel (1990)]. Ultrafiltration The ultrafiltration process consists essentially of the following steps: 0 Separation and cooling of whey (whey is very unstable and cooling re- duces chemical and microbiological changes). 'Clarification/filtration of whey and passing through the ultrafiltra- tion module. 'Forcing of the feed liquor in the ultrafiltration module across the membrane surfaces under pressure until the desired or maximum possible concentration is reached. Ultrafiltration can increase the concentration of milk proteins to 10- 15% solids. 0 Pasteurization of the protein concentrate and further concentration by standard evaporation systems to about 45% total solids. 0 Spray drying by conventional methods [Christensen (1976), Crocco (1975)]. During ultrafiltration the components of whey are fi'actionated as a function of their size and structure, by means of a pressure gradient and a semi-permeable membrane. By this process two fractions are obtained. The mine whiCI brani wate tion t mem mole temp PIES: (199: are c Tabl 14 concentrate or retentate consists of whey components that cannot pass the membrane pores, predominantly proteins, fat globules, suspended solids, min- erals and vitamins bound to proteins, lecithin and enzymes. Since separation is not complete, the product is a WPC which still contains some lactose and minerals dissolved in water. The other fraction is the filtrate or permeate, which consists of whey components smaller than the pore size of the mem- brane such as lactose, unbound minerals, organic acids, non-protein nitrogen, water soluble vitamins and water. A schematic representation of ultrafiltra- tion of whey is demonstrated in Figure 1 [Marshall (1982)]. The ultrafiltration membranes commonly used in dairy processing retain the components with molecular weight of 10,000 or above. Ultrafiltration usually takes place in the temperature range of 50°F-122°F (10°C-50°C) using 1-10 bar (100-1,000 kPa) pressure [Horton et a1. (1972), Roualeyn et al. (1971), Siekiewicz and Riedel (1990)]. The composition changes occurring during the ultrafiltration process are demonstrated in function of volume reduction and concentration ratio in Table 3. oooOooOo- *0..Ooo.*o‘ .0 e e o.*. ' “09'6“"! ° * ° ° 0*.Retmsnuow so 0 e . O '0'. .‘O. ii9-i-mm ....,. ..... :e".-.: :e:*..0..*:.e.....:.....'* :.Permutc '1 - ’. : .° : '. 3* --'.-’.. O’Hara: * me Outflow]: . W810! Figure 1. Schematic representation of the ultrafiltration of whey [Marshall (1982)]. cess c Sage cosity and n the c: the h (1983 have rosivi those fivet IEnt, tanee ShOW' 15 Table 3. Composition changes of whey occurring during ultrafiltra- tion [Babella (1984)]. Volume reduction (%) 0 50 60 80 90 95 Composition (%) in dry basis Whey concentrates Protein 12 22 24 37 52 66 Lactose 79 69 68 56 41 28 Mineral salts 9 9 8 7 7 6 Whey proteins with high purity can be produced in a multi-phase pro- cess called diafiltration. In diafiltration, water is added to the feed in the final stages of the ultrafiltration. The water dilutes the retentate, decreases the vis- cosity and, as it permeates, washes out lactose, non-protein nitrogen (NPN) and minerals. The end result is to increase the purity of the whey protein in the concentrate by the reduction in NPN, lactose and ash, while maintaining the true protein concentration essentially constant [Goldsmith (1981), J elen (1983). Marshall (1982), Muller (1976), Sienkiewicz and Riedel (1990)]. The ultrafiltration membranes do have physical limitations which have contributed to their early lack of acceptance. Extremes of pH, heat or cor- rosive chemicals, for example, will corrode first generation membranes, such as those constructed of cellulose acetate. Moreover, cellulose acetate was sensi— tive to microorganisms and some commonly used disinfectants. To a great ex- . tent, however, these limitations have been overcome. Second generation mem- branes, such as polysulphones or polyamides, offer improved corrosion resis- tance resulting in greater acceptance [Anon (1981), Marshall (1982)]. Table 4 shows the properties of some ultrafiltration membrane materials. by pr Othei assoc incorr tion c impaj trafili the dc eratu tion 17 also a CrObi; ing s ilab'o (1982 (1986 16 Table 4. UF membrane materials and their properties [Anon.(1981)]. Material type pH range Max temp. Chlorine Solvent at pH 7 resistance resistance Cellulose acetate 4.5- 9 55°C Good Poor Polyamide 3.0- 12 80°C Poor Good Polysulphone 0 - 14 80°C Good Good Polyacrylonitrile 2.0- 12 60°C Good Poor Polyfuran 2.0- 12 90°C Poor Good The major problem in the ultrafiltration of whey is membrane fouling by proteins and salts which become concentrated on the membrane surface. Other limitations include prolonged exposure to elevated temperatures with associated problems due to microbial contamination, protein denaturation and incomplete removal of low molecular weight components. Membrane compac- tion causes a serious permeate flux reduction at higher operating pressures. Fouling may result in a short-term decline of flux or in a permanent impairment of the membrane permeability. Pretreatments of liquids to be ul- trafiltered may be employed to increase the rate and/or to change the nature of the deposit formation. Some pretreatments that have been suggested in the lit- erature include clarification by centrifugation or filtration followed by separa- tion for removal of cheese fines and fat that will contribute to fouling and may I also affect the quality of the WPC, beefing for reduction of the viscosity and mi- crobial loads of whey, microfiltration for separation of fat and bacteria by pass- ing through a membrane with pore size of 1.2 pm, pH adjustment, demineral- ization and preconcentration [Marshall (1982), Matthews et at. (1978), Morr (1982), Richter (1983), Sienkiewicz and Riedel (1990), Tamawski and Jelen (1986)]. The ultrafiltration of acid whey results in lower flux rates than sweet 17 whey. Calcium apatite, whey protein structure and their interactions have been implicated. For acid whey pH adjustment from 4.6 to 7.0 is usually ac- complished with N aOH or Ca(OH)2, that yields calcium phosphate, which ab- sorbs fat and protein and is separated by gravity [Sienkiewicz and Riedel (1990)]. Sanitation can be another problem, because membrane systems must be scrupulously cleaned and microbiologically monitored, to avoid blockages, poor flux and membrane fouling may occur. Membranes should not be physi- cally handled, so clean-in-place schemes with high velocity liquid streams are used. Two to four hours of cleaning includes cleaning cycles followed by a san- itation cycle [Keck Membranes, Inc]. - Another disadvantage of membrane systems has been the high capital and membrane replacement costs. Higher capital costs, however, are ofi'set by other advantages. Because they operate at lower temperatures, and no phase change (from liquid to vapor) is involved, membrane systems use less energy than other systems. Thus, the feasibility of membrane filtration is tied to low operational costs [Anon (1981), Sienkiewicz and Riedel (1990)]. Functional properties of whey protein concentrates Whey protein concentrates obtained by ultrafiltration and other mem- brane separation processes are generally up to 90% water soluble in the pH range 3-8. Application of heat during spray drying has little effect on solubili- ty. Pasteurizatibn of the WPC solution can, however, result in a denaturation of up to 20% and a solubility loss in the isoelectric range [Sienkiewicz and Riedel (1990)]. 18 Fat emulsifying capacity is defined as the oil quantity, in grams, which is retained by one gram of protein of the WPC under prescribed conditions. Whey protein concentrates have emulsion capacity values which are worse than those of sodium caseinate. This is due to their comparatively more regu- lar sequence of hydrophobic and hydrophilic groups and their more compact, globular conformation [Muller (197 6), Sienkiewicz and Riedel (1990), Smith (197 6)]. The whipping properties of WPCs are variable, and values ranging from 0 to 680% were found for metaphosphate complex and gel filtration pro- tein concentrates. Whipping ability is the amount of air which is incorporated into a given amount of sample during its churning for a given period of time and is usually expressed as a percentage of the sample volume. Whey protein concentrate solutions with 10% protein can give good whips, but the presence of more than 2% fat can adversely affect whipping, and there is some evidence that whippability is affected by the temperature history of the sample, pH, clarification, calcium level and the addition of such materials as sucrose and hydrolyzed starch [Smith (1976)]. Similarly, the patterns of buffering capacity versus pH for the WPCs are varied. Only those prepared by electrodialysis, ultrafiltration and gel fil- tration were somewhat similar in format with low bufl‘ering capacity at pH 7 .0, which gradually increased in the lower pH regions [Smith (197 6)]. Water-holding capacity of a substance is the grams of water bound to ' 1 gram of dry matter of this substance. The water-holding capacity of WPCs is dependent on the protein concentration, the mineral content and the degree of the denaturation of the proteins. For native and denatured whey proteins the water-holding capacity ranges from 0.5 to 1.2 g H20 per 1 g dry matter and is 19 very low in comparison to both soya protein concentrates and sodium caseinate solutions [Sienkiewicz and Riedel (1990)]. The protein content of the WPC which can be reached by using ultrafil- tration is practically limited to about 80%. At these high protein contents, the functional properties are marred by the retention of fat by the UF membrane, thus causing loss of whipping properties. For this reason, removal of residue . lipids before the ultrafiltration is very important for improving whipping and, perhaps, other functional properties of the WPCs [Marshall (1982), Ryder (1980), Sienkiewicz and Riedel (1990)]. USES OF WHEY IN FROZEN DESSERTS The standards of identity for ice creams do not allow replacement of MSNF by whey solids at a level higher than 25%. Ice cream must have, at least, 2.7% milk-derived protein by weight [CFR (1990b),Weiner (1977)]. A ma- jor reason for use of whey solids in ice creams and other frozen desserts is that it is the least expensive dairy product that can be used in such formulations. A Modified whey products have been successfully used to contribute to the milk solids non-fat content of ice cream and other frozen desserts. Use of modified whey in ice cream is said to eliminate or reduce the effect of some ob- jectionable changes that we encounter in ice cream, such as shrinkage and sandiness. Shrinkage is a term usually applied to contraction of volume of packaged frozen dairy products, a defect caused by the loss of overrun air. The product shrinks from every direction, pulling away from the sides and tap of the container. Sandiness is manifested as a powdery and gritty sensation, which can be perceived on the tongue even after the product has melted. This she: 25, 5 la ice whey Leigh SNF . COD tro S'ubsti't 0f the I 010% 1 ice Cree sOlids. 20 defect is due to the presence of large lactose crystals. Some factors to initiate the development of sandy texture are high lactose content and high and, per- haps, fluctuating temperatures. Dry whey contains a greater amount of lactose (72%) than does NFDM (52%) or other common source of MSNF. This fact has caused some concern that the use of dry whole whey might cause sandiness in frozen dairy desserts. This is the reason why partially delactosed whey has been used in fi-ozen desserts [Anon (1979), Martinez and Speckman (1988)]. Crowe (1960) replaced 50, 75 and 100% of MSNF with dry whole whey in vanilla and strawberry ice cream mix containing 11% solids non-fat (SNF). No preference for the flavor of either control samples or samples containing whey was found. Frazeur ( 1959) found that substitution of dry sweet whey at 25, 50 and 75% level for MSNF did not affect the consumer acceptance of vanil- la ice cream. A taste panel could not differentiate ice cream containing dry whey replacing the MSNF up to 75% from ice cream containing no whey. Leighton (1944) compared five vanilla ice creams containing 8% butterfat, 6.4% SNF and 15% sugar, which had different source of the SNF. In the first ice cream, all SNF came from NFDM, whereas the others contained sweet whey solids to an amount equalling 1, 2, 3 and 4% of the mix. The body and texture of samples containing 1% whey solids were equal or slightly better than the control, whereas the samples containing more whey solids were found inferior than the control. In no case, however, were undesirable flavors noted. Similar Substitution experiments, with mixes of higher fat and SNF content than those of the previous experiments, gave the following optimum whey solids content: i) 10% fat and 8% SNF ice cream 1.6% whey solids; ii) 12% fat and 9.6% SNF ice cream 2.3% whey solids; and iii) 14% fat and 8% SNF ice cream 3% whey solids. Excellent sherbets of 2.6% fat and 2.4% SNF, where 78% of the SNF 21 came fiom whey solids were also produced. These sherbets were not noticeably different fi'om the control samples made entirely with MSNF. Reid with Shaffer (1947) found that excellent chocolate and strawberry ice creams and good vanilla ice cream can be obtained, even when 90.9% of the MSNF are replaced by dehydrated whole whey solids. At this high level of sub- stitution the vanilla flavored ice creams had a slight heat flavor and their tex- ture was ranked as “good” after 30 and 40 days of storage at -14°F (-25.6°C). The substitution levels of whey solids for MSNF tested were 9.1, 18.2, 27.3, 45.5, 63.6 and 90.9%. Vanilla ice cream kept for 4 days in the hardening room and then transferred to cabinets at 5.2°F (-14.9°C) The samples with 63.6 and 90.0% substitution of whey solids for MSNF had slight heat flavor, good texture and excellent body after 10 days of storage, whereas alter 30 and 40 days at 5.2°F (-14.9°C) slight sandiness was reported. Potter and Williams (1949) used dry sweet whey (95% total solids), plain condensed whey (60% total solids), sweetened condensed whey (80% total solids, 40% sugar and 40% whey solids) and fluid whey (6.4% total solids) from Cheddar and Swiss cheese to formulate sherbets containing 4.42-5% whey sol- ids. The finished products possessed fine flavor, body and texture. Among sev- eral advantages to be gained through the use of spray dried whey powder in ice cream reported by Rosenberger and Nielsen (1955) were a smoother body and texture, a higher melting resistance and easier incorporation in mixes than NFDM. The same researchers also reported that the flavor of the ice cream may be slightly inferior, when whey powder is used. Frazeur (1967) substituted electrodialyzed dry sweet whey and excel- lent and average flavor dry wheys for MSNF in ice cream, ice milk, soft-serve ice milk and shake mixes at 25% level and in sherbet at 64.7% level. He found that the flavor of ice cream, ice milk, soft-serve ice milk and milk shake sam- 22 ples which contained electrodialyzed dry whey was equal to the flavor of control samples. However, statistically significant differences between the flavor scores of control samples and samples containing the dried whey of average fla- vor were found. No significant difl‘erence in body and texture occurred due to the use of any of the whey products. For the sherbet samples the flavor, body and texture was significantly improved by the presence of all the whey prod- ucts. StOrage of sherbet samples at 3°F (~16.1°C) for nine weeks did not change the relative differences among flavor, body and texture scores for all sherbet samples. Storage did, however, decrease, flavor, body and texture scores for all sherbet samples. In a sensory study of the previous mentioned frozen desserts conduct- ed by Frazeur and Harrington (1967) consumers responded to orange sherbet in a manner which was not observed with any of the other fiozen dairy products that were tested. They indicated significant overall and flavor preferences for samples which contained some form of dry whey. However, the smoothness of the control sherbet samples was preferred. The results indicated that the use of electrodialyzed dry whey is to be preferred in sherbets and in soft-serve ice milk, an excellent flavor quality dry whey should be used in sherbet, should not be used in ice cream and can be used in ice milk, soft-serve ice milk or milk- shake mix and an average flavor quality dry whey probably should not be used- in any frozen dairy dessert, with the single exception of sherbets. Arnold et al. (1976) compared the effect of using various levels of dried sweet whey and partially delactosed whey solids in ice cream formulations. Characteristics that were evaluated included flavor and texture of the finished ice cream. Three levels of replacement wereused, 20, 35 and 50%. All levels of replacement of MSNF by either type of whey had little effect on samples held no longer than four weeks. The results of this study indicated that use of up to 35% r mixfc pam's used: total s ent 1e ferem ingre ere fereni 5.5% creas. Wt 54 sured table 5.5% for sc 00w Con d. that Whex lated and 1 feet ( the Q K. 23 35% replacement MSNF with dried sweet whey may be acceptable in ice cream mix formulations. To avoid adversely afl‘ecting flavor and texture of ice cream, partially delactosed whey powder with a high mineral content would need to be used at levels lower than 20% replacement of MSNF. In a study conducted by Guy (1978), vanilla-flavored ice creams of 38% total solids containing 12% fat and 0.14% stabilizer were prepared with differ- ent levels of sweet whey (0-11%) with either 67 or 79% hydrolyzed lactose, dif- ferent levels of sugar (10- 15%) and different levels of MSNF (541%), the three ingredients totaling 26%. The initial hedonic flavor and texture scores of ice creams containing up to 5.5% of either of the wheys were not significantly dif- ferent from those of the controls. Increasing both types of wheys solids above 5.5% significantly decreased mix viscosity and hedonic flavor scores and in- creased the saltiness of the ice cream. Texture scores fell less rapidly than fla- vor scores, although whey contributed to a softer-bodied ice cream, as mea- sured by compressibility tests. The sweetness of all test samples were compa- rable to the control. Heat-shock stabilities of the ice creams containing up to 5.5% whey were good; they were poorer for those with higher levels of whey and for some of the controls. Dry sweet whey has often been used successfully in ice cream at low concentrations, but WPC show the most promise in replacing NFDM. A study, conducted by Huse et al. (1984), evaluated the limits of the proportion of WPC that could be used in ice cream manufacture. The use of WPC rather than whey powder would allow the maintenance of high protein levels in the formu- lated ice cream. The level of replacement of MSNF with whey solids was 0, 50 and 100%. Fifty percent replacement of MSNF with whey solids had little ef- fect on the sensory qualities of the ice cream. The use of only WPC to supply the SNF gave a poorer textured ice cream with some increase in iciness and sig- 24 nificantly decreased smoothness, creaminess and fullness of flavor. The flavor of this ice cream was also very flat and more cooked due to high proportion of heat labile whey protein. Samples containing whey solids had slightly better resistance to heat shock with no increase in iciness score and only a slight de- crease in smoothness alter the applied heat shock treatment. In a study conducted by Parsons et al. (1985), ice cream was made from mixes in which 100% or 50% of the MSNF was provided by WPC, WPC and dried sweet whey (DSW) or DSW and sodium caseinate (CAS). Trained panelists found no significant differences in flavor, body and texture among the ice creams, but in a 14-week evaluation by 52 randomly selected families the DSW-GAS ice cream was found inferior in flavor compaerd to the other prod- ucts. The consumer study rated the WPC and WPC/DSW ice creams as equal or slightly better than the MSNF control ice cream at both 50 and 100% re- placement. The panelists made various comments about the samples contain- ing WPC, such as being “creamier” and “smoother” tasting. They also indicated that the ice cream samples containing the whey blend were sweeter and very similar in taste to soft serve ice cream. In a preliminary study conducted by Coder and Parsons (1979) similar results were obtained. Flavor problems associated with the use of Cheddar cheese whey in the formulation of ice cream mix were investigated by Bodyfelt et al.(1979). An ice cream model system was used to study the effects of varying whey quality and quantity and extent of heat processing on the flavor profile of the final product. Whey “fingerprint” compounds were identified by headspace gas liquid chro- matography/mass spectroscopy (GLC/MS) analysis of varied quality whey pow- ders. The chemical compounds that appeared most representative of the heat- ed, stale ofi-flavor of dried whey included four difi‘erent pyrazines, m-pentanol, 25 dimethyltrisulfide, 2-furfural, benzaldehyde, 2-furfuryl alcohol and dimethyl- sulfone. The sales potential of whey products is not only determined by the whey processing technique, but also by the quality of the whey available. Gen- erally, the value of the whey is increased in relationship to the decreased con- centration of lactic acid in the whey. Whey from Cottage cheese is difficult to process and generally is disposed of without value adding processing, whereas sweet whey from Swiss or Emmanthaler cheese is in special demand [Chris- tensen (1976)]. Watrous et al.(1991) found that the use of concentrated acid whey, both neutralized and unneutralized, to contribute 8.9% by weight in the serum sol- ids in vanilla flavored ice cream resulted in a product which could not be differ- entiated finm an ice cream product containing no whey solids and an ice cream product containing 24.7% sweet whey solids by weight of the serum solids. The use of concentrated neutralized and unneutralized acid whey to contribute 26.6% by weight of the serum solids in vanilla flavored ice cream resulted in a product of poor texture, overrun and color. Experiments with neutralized acid whey found that it can be used to contribute 17 .7% by weight of the serum sol- ids, without impairment of any undesirable sensory properties to the ice cream. In a work that has been done by Patel and Harper (1977), acid whey was concentrated to about 20% total solids by reverse osmosis and was used to replace 10-25% of the MSNF in ice cream with basic composition 10% milk fat, 12.5% MSNF, 14% sucrose and 0.3% stabilizer and emulsifier. Compared to the control ice cream, the experimental sample with 10- 15% replacement had higher viscosity after processing and after 20 and 44 hours aging and exhibited higher pseudoplasticity. Partial replacement of MSNF did not affect freezing time, but resistance to melting was increased, with the 20% replacement show- 26 ing the greatest resistance. N o statistically significant differences were ob- tained in organoleptic evaluation. Igoe et al. (1973) used Cottage cheese whey in ice cream mixes and found it unacceptable when more than 1% of the SNF were provided by acid whey. In this study, the SNF were standardized at 11.25% by weight for all mixes. In the control mix, these solids were derived from skim milk, whereas in the other mixes whey solids were substituted for varying amounts of skim milk solids, so that the 11.25% level was maintained. 1%, 2%, 2.8% and 3% by weight of whey solids were incorporated into the ice cream, representing sub- stitution levels of 8.9%, 17.7%, 24.7% and 26.7% respectively. The whey ingre- dient was fresh Cottage cheese whey concentrated to 29% solids in a vacuum pan. Concentrated neutralized acid whey and sweet whey were also used for comparison. The sensory panel could not distinguish between ice cream con- taining 1% acid whey solids and the control. At the 2% level, however, ice cream with acid whey solids was easily differentiated with a strong preference shown for the control. At the 3% level the flavor and texture properties were so obviously different that sensory testing was not necessary. The results also showed that ice cream with 2% neutralized acid whey solids was much pre- ferred over that containing 2% of unneutralized acid whey, but the authors did not recommend the use of neutralized acid whey at levels higher than 2%. Ice cream with 2% acid whey was readily distinguished from ice cream containing 2.8% by weight of sweet whey, but when the comparison involved neutralized acid whey, the difi'erences found were not statistically significant. Potter and Williams (1949) made sherbet using Cottage cheese whey containing 6.4% total solids, which resulted in a product with 4.42% whey sol- ids. The researchers found that the final product possessed a smooth body and texture and was more refreshing than the sherbets made with solids from milk 27 or ice cream mix, and that no citric acid needed to be added to the mix when Cottage cheese whey was used. Hekmati and Bradley (1979) used Cottage cheese whey in sherbet as a replacement for water. Sherbet formulations were prepared containing as high as 65.4% whey as a diluent. Flavor, body and texture were rated excellent in all sherbets evaluated. However, the sample prepared with the highest level of acid whey (65.4%) showed slight masking of the pineapple flavor of the sher- bets. Direct-set Cottage cheese whey was used by Demott and Sanders (1980) for the manufacturing of sherbet. Three sherbet mixes containing (i) 28.5% milk and 17.2% skim milk (control), (ii) 44.7% whey and 8.3% ice cream mix and (iii) 46.1% whey and 9.5% half-and-half were compared. The same amount of citric acid was added to all mixes before freezing. The last two mixes had higher titratable acidity (0.86 and 0.85% versus 0.61% for the control), slightly lower pH (3.5 and 3.5 versus 4.0 for the control) and slightly higher total solids (31.7 and 32.0 versus 31.1%). An expert panel was unable to detect any defect attributable to the whey in sherbet samples prepared by mixes (ii) and (iii) and an untrained panel detected no differences among the sherbets. The conclusion of this study was that direct-set whey may be used to replace some of the water and milk solids in sherbets without adversely affecting their flavor. Demott noted in one of his other studies that whey from direct acidifi- cation process does not have the “whey taint” associated with cultured whey. It has also been reported that making whey products immediately after the whey is separated from the curd eliminates deterioration caused during storage [McGugan et al. (1979)]. Chapter 3 EXPERIMENTAL PROCEDURE PREPARATION OF WHEY PROTEIN CONCENTRATES Two hundred forty gallons (908.4 L) of direct-set Cottage cheese whey were supplied by Country Fresh Inc. [Grand Rapids, MI] and were processed the same day they were produced to minimize spoilage problems. The whey was ultrafiltered and diafiltered in the Kock Membranes S-l ultrafiltration pi- lot system [Wilmington, MA] after the necessary cleaning and sanitation of the system. The temperature of the system during the ultrafiltration/diafiltration was kept at 130i2°F (54:1:1°C), and the pressure drop between the inlet and outlet valve was 101:1 psi (689516.90 kPa). During the operation of the system the volumetric rate of the permeate was determined as Q=0.413ga1/min (1.563L/min). When the whey was concentrated to about 25 gal (94.63 L), 100 gal (378.5 L) of soft water were added to the tank for diafiltration in order to de- crease the lactose and mineral levels of the product. The operation was inter- rupted when the whey was at the level of 25 gal in the tank. The concentration factor of the process, which is the ratio of the volume of the whey before and after ultrafiltration was CF=250/25=10 and the total weight of the concentrat- ed product was 177.5 lbs (80.51kg). The product was collected in marked plas- 28 29 tic bags with caps and kept in the freezer at temperature -13°F (-25°C), to min- imize spoilage. TOTAL SOLIDS DETERMINATION ' The total solids of the original acid whey, the WPC, the difl'erent or- ange sherbets and sherbet mixes were determined by the Mojonnier method [AOAC (1990)]. FAT DETERMINATION The fat content of the acid whey, the WPC, the cream and the sherbet mixes was determined by the Mojonnier method [AOAC (1990)] with some modification. For the whey protein concentrate and the sherbet mixes no rec- ommendation for the amount of reagents was found in the literature [Newland- er and Atherton (1964)]. Thus the whey protein concentrate, because of its acid nature, was treated as acid whey and 3ml of ammonium hydroxide were used, and the sherbet mixes were treated as ice cream with fat content similar to that of fresh milk. The samples of sherbet mix examined were 10 g and the amount of reagents added was: 5 ml water, 3 ml ammonium hydroxide, 10 ml alcohol, 25 ml ethyl ether, 25 ml petroleum ether for the first extraction and 5 ml alco- hol, 15 ml ethyl ether and 15 ml petroleum ether for the second extraction. 30 TOTAL NITROGEN DETERMINATION Total nitrogen was determined by the Kieldahl nitrogen method [AOAC (1990)]. The nitrogen determination system used in this study was composed of the Tecator 40/1016 digester, the Buchi 322 distillation unit, the Buchi 342 control unit, the Metrohm 614 impulsomat, the Metrohm 632 pHme- ter and the Metrohm 655 Multi-Dosimat unit [Buchi Laboratories, Flawil, Switzerland]. From the volume of HCl consumed for each sample and the blank the % total nitrogen was calculated by using the following equation: (HCls-HClb) = Sample weight 70 TN xAxNormalityHCl where: I-ICl8 = volume of HCl consumed for each sample’(m1) HClb = volume HCl consumed for the blank (m1) Sample weight = weight of tested sample (g) A = 1.4007 (g/mol). NON-PROTEIN NITROGEN DETERMINATION The preparation of the samples for the non-protein nitrogen (NPN) de- termination is described by Partridge (1983). It was assumed that all the pro- teins are insoluble in 12% TCA and therefore, NPN was determined as soluble nitrogen in 12% TCA filtrate prepared from each sample. 31 The Sorvall RC-5B Refiigerated Superspeed Centrifuge (Du Pont 00., Wilmington, DE) was used in this determination and the samples were subject- ed to 13,000 RPM (Relative Centrifugal Force: 20,000xg) [Cooper (1977)]. pH DETERMINATION The pH of the original acid whey, the WPC, the sherbet mixes and the sherbets was determined using the Coming pH/ion meter 145 with the Corn- ing semi-micro combination electrode [Medfield, MA]. The pH meter was cal- ibrated with Bufi'ar pH 7.0 and pH 4.0 buffer stande solutions (Mallinckrodt Inc., Paris, KY). TITRATABLE Acmrrr The titratable acidity of the WPC, the sherbet mixes and the sherbets was determined with the Nafis apparatus (Meyer-Blanke Co., St. Louis, MO) described by AOAC (1990). PREPARATION OF SHERBET MIXES Four sherbet mixes (Table 5) were prepared, one control, containing no whey (sherbet mix A), one with 25% substitution of whey solids for MSNF (sherbet mix B), one with 50% substitution (sherbet mix C) and one with 75% substitution (sherbet mix D). All mixes contained 1.5% milkfat, 3.4% solids 32 non-fat (SNF), 20.0% sucrose, 9.0% corn syrup solids (CSS), 0.3% stabilizer/ emulsifier (STEM) and 34.2% total solids. The only difl'erence they had was the source of the SNF. Table 5. Formulation of the sherbet mixes. WWWW 1.5% fat 1.5% fat 1.5% fat 1.5% fat 3.4% MSNF 2.55% MSNF 1.7% MSNF 0.85% MSNF 20% sucrose 20% sucrose 20% sucrose 20% sucrose 9% CSS 9% CSS 9% CSS 9% CSS 0.3% STEM 0.3% STEM 0.3% STEM 0.3% STEM 0% whey solids 0.85% whey solids 1.7% whey solids 2.55% whey solids The ingredients used in the sherbet mixes are presented in Table 6. The calculation of the required amount ofWPC for each sherbet mix was based on the solids non-fat (SNF) of these ingredients. The solids non-fat of the cream were calculated from its total solids and its fat content. The stabilizer/emulsi- fier (STEM) used in this study, known by the brand name Kontrol [German- town Mfg., Broomall, PA] contains mono- and diglycerides, cellulose gum, guar gum, polysorbate 80, carrageenan and sodium silico-gluminate. The CSS used, known by the brand name Maizo [American Maize Prod. Comp., Hammond, IN] was a 42 Dextrose equivalent (DE) product. Table 7 shows the amount of each ingredient used for the preparation of 80 lb (36.29 kg) of sherbet mix A, B, C and D. All ingredients were mixed together in a 10 gal (37.85 L) container and were pasteurized with continuous stirring at 175°F (795°C) for 5 min. Then the mixes were homogenized at 1500 and 500 psi (10,3425 and 3,447.5 kPa) for 33 the first and second stage, respectively, in a homogenizer made by Manton- Gaulin Mfg. Co. Inc., type 75K (Everett, MA). The mixes were then collected in plastic bags, cooled down to 40°F (4.44°C) with ice and stored overnight in a cooler at 40°F (4.44°C). Table 6. Ingredients used in the formulation of sherbet mixes. Increments WW Cream 36.601026 fat, 418010.26 total solids NFDMS 4.6610.09 moisture WPC 01110.01 fat, 55310.10 total solids Granulated sugar Corn syrup solids 42 DE 44910.09 moisture Stabilizer/emulsifier (STEM) 4.7910.01 moisture Tap water Table 7. Amount of ingredients used for 36.29 kg of sherbet mix A, B, C and D. W W W W 1.49 kg cream 1.47 kg cream 1.45 kg cream 1.44 kg cream 1.21 kg NFDM 0.89 kg NFDM 0.57 kg NFDM 0.24 kg NFDM no WPC 5.69 kg WPC 11.38 kg WPC 17.07 kg WPC 7.26 kg sucrose 7 .26 kg sucrose 7.26 kg sucrose 7.26 kg sucrose 3.42 kg CSS 3.42 kg CSS 3.42 kg CSS 3.42 kg CSS 0.11 kg STEM 0.11 kg STEM 0.11 kg STEM 0.11 kg STEM 22.80 kg water 17.45 kg water 12.10 kg water 6.75 kg water FREEZING OF THE SHERBET MIXES The sherbet mixes were frozen in the Gelmark model 160 pilot plant continuous freezer [Alfa-Laval, Hoyer, Italy] with 43% overrun. Before fi'eez- 34 ing, 75.8 ml Bloomfield orange sherbet base [Kraus & Co. Inc., Oak Park, MI], 12 ml 50% w/w citric acid solution and 8 m1 orange food color [Kraus & Co. Inc., Oak Park, M1] were added to 10 lbs (4.536 kg) of each mix. The mixes were frozen and the sherbet was packaged in paperboard 16T5 pint containers (Ne- style, Sealright Co. Inc., Fulton, NY) and hardened at -13°F (-25°C). After a week, the sherbets were transferred and kept in another freezer at 0°F (~17.8°C) to approximate the temperature of home freezers. OVERRUN DETERMINATION The overrun of the different sherbets was calculated during the freez- ing. The equation used was the following: % OR = ng: Sh” x 100 Sher where: % OR = % overrun (dimensionless) Wm = mass of a certain volume of mix (g) Wu," = mass of the same volume of sherbet (g). 35 DETERMINATION OF THE RHEOLOGICAL PROPERTIES OF THE SHERBET MIXES The rheological properties of the sherbet mixes were tested in the Haake RV12 concentric cylinder viscometer [Haake Buchler Instruments, Sad- dle Brook, NJ] using the MV cup with the MV-I sensor and a M500 measuring head. The sherbet mixes were tested two days after they were prepared. Dur- ing this period, they were aged at 40°F (4.44°C). Two tests were run on each mix. The time-dependency of the rheological properties of the samples was tested by the first test , while the best rheological model for the description of the mixes was developed for the time-independent samples by the second test. Each test was performed in triplicate. A material has a time-dependent behavior, if when subjected to a con- stant shear rate, the shear stress increases (for rheopectic materials) or de- creases (for tbixotropic materials) over time. The time-dependent thickening of the fluid is known as rheopexy, whereas the time-dependent thinning as thixotropy. The viscometer was set at 40°F (4.44°C) by using the Haake C water bath [Haake Buchler Instruments, Saddle Brook, NJ]. The reason for this set- ting was that the sherbet mixes were kept in the cooler at 40°F (4.44°C). For the time-dependency test, the viscometer was set up to rotate for 10 min at the constant angular velocity of 100 min'l. The torque was measured every 12 sec (the 10 min period was divided into 50 intervals) and the diagram of torque ver- sus time was obtained. For the second test, the angular velocity was gradually increased from 0 to 500 min'l, by using a Haake PG 142 programmer [Haake Buchler Instruments, Saddle Brook, NJ]. The angular velocity was increased 36 in small intervals for low values and bigger intervals for high values. For each angular velocity value the torque applied to the sample was recorded by the computer unit of the viscometer. The shear stress was calculated based on standard procedures and the shear rate based on the Krieger method [Stefi‘e (1992), Whorlow (1979)]. The shear stress versus shear rate diagram was ob- tained in addition to the raw data table and the statistical analysis of the data. The properties of the mix samples were measured over shear rate range up to 1100 s'l. MELTING RESISTANCE TESTS After the paperboard container was pealed ofl', one pint of each sherbet was placed into a large funnel inserted in a graduated cylinder. The melting took place in a temperature controlled incubator at 100.4¢2°F (3811°C). A Rapid-Flo 6.5 in single gauze faced milk filter (J ohnson-J ohnson, Chicago, IL) was used in the funnel, which retained the foam, letting only the liquid sherbet pass. The volume of melted product was recorded every 5 min, and a graph 1 was obtained by plotting the average volume of sherbet collected versus time. Two replications were performed on each sherbet. SENSORY EVALUATION The objective of this study was to measure the overall flavor and tex- ture acceptability of the sherbets by consumers and compare some of their sen- sory attributes. The final products were tested for flavor and texture accept- 37 ability using a 9-point hedonic scale (Appendix C, Table 0.1), where 1 = “dislike extremely” and 9 = “like extremely”. For sweetness, creaminess and iciness a multiple paired comparison test was used. The acceptance tests were conduct- ed 8 days after the sherbet mixes were frozen, whereas the iciness, sweetness and creaminess were tested 9, 10 and 11 days, respectively after the sherbets were flown. The questionnaires were developed and the trays with the sam- ples were set the day before the test. The panels for the flavor and the texture acceptance test consisted of 48 untrained members of Michigan State Univer- sity (MSU) including faculty and students. For the sweetness, creaminess and iciness tests trained panelists were used. The sweetness panel consisted of 5 people (4 females and 1 male), the creaminess panel of 5 people (4 females and 1 male) and the iciness panel of 6 people (4 females and 2 males). Each test was conducted individually, in order to eliminate any biases. The tests were performed in a sensory evaluation laboratory. The pan- elists were sitting in individual booths, which were lighted properly with bright white light. The room was free of odors, reasonably quiet and comfortably warm. The efforts made to reduce experimental error during the tests, in ad- dition to the above mentioned controlled atmosphere conditions, were: precise instructions for the panelists, avoidance of bias by cleansing the palate with water before and after tasting each sample, identical preparation of the sam- ples and balanced order of presentation [Larmond (1987), Ott (1990)]. Each tray had the following items: ' One sample for the acceptance tests (sherbet A, or B or C or D) and two samples for the multiple paired comparison test. ' One questionnaire (Tables C.1-C.3 in Appendix C). ' One pencil. ' One cup of tap water at room temperature. 38 ' One napkin. ' One expectoration cup. The samples were served in 1 oz. plastic souflle cups, type P100 (Solo Cup 00., Urbana, IL) covered with plastic snap-on lids, No. PLl (Solo Cup 00., Urbana, IL). The cups were marked with a random three digit number. They were filled with a scoop of sherbet, put on trays and placed in the fi'eezer the day before the test [Larmond (1987), Meilgaard et al. (1987a), Stone and Side] (1985)]. The panelists were instructed to rinse their mouth with water after tasting each sample, in order to avoid any influence from the previously sam- pled sherbet. For the same reason, a monadic sequential presentation order was used for the acceptability tests, where each sherbet was presented to the panelists alone accompanied by a new questionnaire. The panelists were also asked to write comments. All 24 possible combinations of sample presentation were used to eliminate the order effect. With the term “order effect” the efl'ect on the results of the presentation order of the samples to each panelist is meant. The paired comparison attribute tests were performed in triplicate. All replications were performed the same day to eliminate possible intensity changes of the attributes over time. The Schefl‘e multiple paired comparison test was used. In this test the panelists were asked to indicate the size of the difi‘erence detected. All samples were compared with every other sample (6 possible pairs for the 4 sherbets), and all pairs were presented to all judges. Half of them tasted one sample of each pair first, the other half tasted the same sample second (balanced design). The order of presentation of the 6 pairs was randomized for each judge and a 7-point scale (Appendix C, Table C.3) was used [Lamond (1987), Ott (1990)]. 39 The hypotheses, and the worksheets for all tests are presented in Ap- pendices D and E, respectively. STORAGE STABILITY The storage stability of frozen dairy products can be reflected as changes in the overall flavor or texture. In order to determine the stability of the products during storage at 0°F (-17.8°C), which is the temperature of a commercial type retail fi-eezer, the sherbets were evaluated for their flavor acceptance after 31 and 122 days of storage and for their texture acceptance after 31 days of stor- age. Forty eight untrained panelists, students and faculty of MSU, were used for each test. The four different sherbets were also compared for iciness by the expert panel after 32 days of storage. HEAT-SHOCK STABILITY 'IVvo weeks alter the freezing of the sherbets, 15 pints of each sherbet sample were taken out of the freezer at 0°F (-17.8°C) and left at room temper store at 7812°F (25.611°C) for 30 min. Then the sherbets were put back to the freezer. The same procedure was repeated for 10 consequent days. A week af- ter this treatment, an evaluation of the flavor and texture acceptance and ici- ness of sherbets was conducted, from the results of which the heat-shock sta- bility of the sherbet samples was compared. 40 PANEL TRAINING Trained panelists were used for the iciness, creaminess and sweetness sensory attribute tests. These attributes are more dificult to evaluate, because everyone has a difi'erent perception about them. The purpose of the training procedure was to create a sensory panel for each attribute, in which each pan- elist was trained to recognize certain characteristics of the frozen dessert, such as iciness, creaminess or sweetness. Twenty four people (students and faculty) participated in this training. The steps of the training procedure follow: 0 Explain and define which attributes were of interest. 0 Prepare sherbets difl'erent in iciness, creaminess or sweetness, by using different amounts of stabilizer/emulsifier, fat or sweetener, re- spectively. 0 Develop a series of sensory tests, to discern the ability of the panelists to distinguish difl‘erences. Testing started with large differences in iciness, creaminess and sweetness among samples. As testing pro- gressed, differences among samples within subsequent trials were of less magnitude than previous trial. 0 At the end of each test, the panelists were told the composition and/ or treatment of the samples they had been tasting, and the characte- ristics of the samples being tested for were emphasized. The pane- lists could ask questions about the samples or the specific sensory test used and could retaste the samples, in order to be able to recog- nize the above mentioned characteristics. 0 At the end of the training period, which included 7 tests for each at- 41 tribute, the panelists who gave the most correct answers (at least 5 out of 7) and were able to find high and low differences in iciness, creaminess or sweetness became the members of the three trained panels. The iciness panel was composed of 6 panelists, whereas the creaminess and sweetness panels consisted of 5 panelists each. The sensory tests that are usually used for panel training are the tri- angle and the duo-trio test. In this project two more tests were used, the paired comparison and the ranln'ng test. The duo-trio test was used alone for the training in sweetness, whereas the others were used for the training in iciness and creaminess. In the triangle test, the panelists tasted three samples, two of which were identical, and were asked to identify the odd sample [Larmond (1987 )1. In this study the panelists were also asked at the end of the test, which sample was the iciest (or creamiest) in order to make sure that they could not only find the odd sample, but they were also able to sense the iciest (or creamiest) sample. The possible combinations of the presentation order of the three sam- ples were six and they were used four times, because the initial panel consisted of 24 panelists. Each sample was presented to the panelist in a cup with a three digit code number written on it. Two code numbers for each sample were selected, so that each tray had three differently coded samples, although two of them were identical. A paired comparison test was used when a triangle test could not be performed or when a lower dificulty level for a specific test was desired. The panelists were given two coded samples and were asked to identify the iciest (or creamiest). There were two possible order presentations of the samples (A- B, B—A) [Larmond (1987)]. 42 In the ranking test three coded different samples were given to the panelists who were asked to rank them from iciest to least icy (or fiom cream- iest to least creamy). In the duo-trio test, three samples were presented to each panelist. One was labeled R (reference) and the others were coded. One of the coded samples was identical to R. Each panelist was asked to identify which of the two samples was difl‘erent from R [Larmond (1987), Meilgaard et al. (1987 a)]. The panelists were also asked to identify the sweeter sample at the end of each test. The same samples were sometimes served to the panelists more than once (in different days), in order to check their ability to differentiate between trials. All sensory tests were performed under red light conditions, to allow the panelists to concentrate on the attribute in question, rather than appear- ance. The final form and schedule of each test were developed after the previ- ous test was over and its results were studied. Training the iciness panel Iciness was defined as the sensation that the ice crystals left in the mouth. The number and size of the ice crystals influence the iciness of the sam- ple. The larger number and size of ice crystals, the icier the product is. The composition and treatment of the samples used for the panel train- ing in iciness are shown in Table 8. Three different stabilization/emulsification systems (STEM) were used in order to obtain difl‘erences in iciness. The Kon- trol STEM has already been discussed. The Dariloid 300 [Kelco Merck Co., Inc., Chicago, IL] was only stabilizer containing gum guar, zhanthan gum and carrageenan. The Rhicoid 200 [Kelco Merck Co., Inc., Chicago, IL] was a 43 Table 8. Composition and treatments of sherbets used for the iciness panel training. Sample Qammsitinn Remnant vanilla ice cream 10.5% fat, 10% MSNF, 12.5%sucrose, homemade 2.5% CSS, 0.25% Kontrol commercial vanilla as above continuous ice cream freezer SH1 1.5% fat, 3.72% MSNF, 20% sucrose, batch freezer 9% CSS, 0.5% Kontrol SH2 1.5% fat, 3.72% MSNF, 20% sucrose, batch freezer 9% CSS, no STEM. heat shock SH3 as SH1 home freezer SH4 1.5% fat, 3.72% MSNF, 20% sucrose, batch freezer 9% CSS, 0.1% Dariloid 300 SH5 1.5% fat, 3.72% MSNF, 20% sucrose, batch freezer 9% CSS, 0.2% Dariloid 200,0.2% Kontrol SH6 1.5% fat, 3.72% MSNF, 20% sucrose, batch freezer 13% CSS, 0.5% Rhicoid STEM system containing mono- and diglycerides, gum guar, zhanthan gum and carrageenan. The recommended amounts of these STEM for use in sher- bets were: for Kontrol 0.30-0.50%, for Rhicoid 200 and Dariloid 300 less than 0.30%. Table 9 shows the samples compared, the sensory tests performed and the percentage of the correct answers on each test. In the first test, the homemade vanilla ice cream was compared to com- mercial vanilla ice cream. The difference in iciness between these samples was high, because the homemade ice cream was whipped by hand resulting in slow freezing, whereas the commercially prepared product was flown quickly in a continuous fi'eezer. Therefore, large ice crystals were developed in the home- made ice cream and they were very easily detected. 44 Table 9. Tests performed for the iciness panel training. Test! Somme: Wat W W 1 home made vanilla triangle test 95.5 ice cream & commercial vanilla ice cream 2 SH1-SH2 paired comparison 92 3 SH1-SH3 paired comparison 95.5 4 SH4-SH5 triangle 33.3 5 SH5-SH6 triangle 43.5 6 SH4-SH5 triangle 52.2 7 SH1-SH4-SH6 ranking 54.6 The difi'erences of the samples in the second test was that SH2 con- tained no STEM and it was heat shocked (left out of the freezer at room tem- perature for 30 min for ten consequent days), whereas SH1 contained STEM and it was not heat shocked. The treatment and the composition of SH2 re- sulted in formation of bigger ice crystals in this sample than in SH1. SH1 was compared to SH3 in the third test. Both sherbets had the same composition, but the first was frozen in a laboratory batch freezer, where- as the second in a home freezer, in which bigger ice crystals were developed. The difference between SH4 and SH5 used in the forth test was the type and amount of STEM used. SH4 was made with 0.1% Dariloid 300, whereas'SH5 with the combination of 0.2% Dariloid 300 and 0.2% Kontrol, so SH5 was better stabilized and developed smaller ice crystals and therefore smoother texture. . The difi‘erences between SH5 and SH6 used in the fifth test was the type and amount of STEM and the amount of CSS used. SH6 contained 13% CSS, whereas SH5 contained 9%, and this ingredient when added at higher 45 levels gives smoother texture to the final product. SH6 was also overstabilized, since Rhicoid 200 was used at almost double concentration than the recom- mended. This test was rather difficult, mainly because SH5 was also a very smooth product. The sixth test was a repetition of test 4. The last test consisted of three samples already tested, SH1, SH4 and SH6. Both SH1 and SH6 were very smooth products and contained 0.5% STEM, however SH6 contains Rhicoid 200, which is not used at higher than 0.3% level. Training the creaminess panel Creaminess was defined as the coating of the mouth left after the sam- ple has been swallowed. The more the sample coats the mouth after swallowed, the creamier it is. The creaminess of a sample is very much influenced by its fat content. Although the samples compared in the first test (Table 11), a premium commercial vanilla ice cream and a vanilla ice milk were quite different in fat content (Table 10), only 50% of the panelists found the premium vanilla ice cream creamier, which was considered as the right response. Alter the test the , panelists tasted some more of each sample and tried to sense the difi'erence in creaminess. A paired comparison test was used, because the samples were quite different in flavor, and the odd sample in a triangle test would be easily recognized. . The samples $111 and SH8 that were used in the second test had dif- ferent fat content. SH8 with 3% fat was very creamy, though above the legal maximum 2% for fat content [CFR (1990c)], whereas SH1 was a legal sherbet. 46 In the third test, SH8 and SH7, having 3 and 2% fat, respectively were com- pared in a triangle test. Although the 1% difference in fat is big, both sherbets were creamier than the usual commercial sherbets, so the panelists found this test dificult. Test 4 was found much easier, although the fat difference be- tween the sample was only 0.5%, because SH7 was richer than the usual sher- bet. Test 5 was a repetition of test 3, but this time an easier sensory test was used (paired comparison), in order to check how many panelists were able to Table 10. Composition of the samples used for the creaminess panel training. Sunnis Qcmmsifion premium vanilla ice cream more than 10% fat (exact composition unknown) . ice milk 4% fat (exact composition unknown) SH1 1.5% fat, 3.72% MSNF, 20% sucrose, 9% CSS, 0.5% Kontrol SH7 2% fat, 3.72% MSNF, 20% sucrose, 9% CSS, 0.5% Kontrol . SH8 3% fat, 3.72% MSNF, 20% sucrose, 9% CSS, 0.5% Kontrol Table 11. Tests performed for the creaminess panel traimng. Test! Salaries W W W 1 premium vanilla ice paired comparison 50 cream & ice milk 2 SH1-SH8 paired comparison 100 3 SH7-SH8 triangle 45 4 SH1-SH7 triangle 76 5 SH7-SH8 paired comparison 75 6 SH7-SH8 triangle 34.8 7 SH1-SH7-SH8 ranking 31 47 find the creamiest sample between the 2 and 3% fat sherbets. The sixth test was an exact repetition of test 3. For the last test SH1, SH8 and SH7 were com- pared. The dificulty of this test lied on the small difference of the creaminess of SH7 and SH8 and the dificulty of the ranldng test itself. Training the sweetness panel Sweetness was defined as the taste stimulated by sucrose and other sugars, such as fi-uctose, glucose, etc. and by other sweet substances, such as saccharin, aspartame and acesulfam K [Meilgaard et al. (1987b)]. The first two tests (Table 12, 13) were easy, because the difference in sugar composition between SH9~SH12 and SH9-SH15 was 5%, which is considered a big differ- ence. Test 3 was also easy (4% difference in sugar content between SH11- SH14). Tests 4 and 5 were harder and the last two tests were the most dificult, because there was only 1% difl‘erence in sugar content between SH9, SH10 and SH9, SH13. Table 12. Composition of the samples used for the sweetness panel traimng. Bennie Comm ° SH9 1.5% fat, 3.4% MSNF, 20% sucrose, 9% CSS, 0.3% Kontrol SH10 1.5% fat, 3.4% MSNF, 21% sucrose, 9% CSS, 0.3% Kontrol SHll 1.5% fat, 3.4% MSNF, 22% sucrose, 9% CSS, 0.3% Kontrol SH12 1.5% fat, 3.4% MSNF, 25% sucrose, 9% CSS, 0.3% Kontrol SH13 1.5% fat, 3.4% MSNF, 19% sucrose, 9% CSS, 0.3% Kontrol SH14 1.5% fat, 3.4% MSNF, 18% sucrose, 9% CSS, 0.3% Kontrol SH15 1.5% fat, 3.4% MSNF, 15% sucrose, 9% CSS, 0.3% Kontrol 48 Table 13. Tests performed for the sweetness panel training. lest! Smlcs W Mamet W 1 SH9-SH 12 duo-trio 75 2 SH9-SH 15 duo-trio 100 3 SH1 1-SH 14 duo-trio 66.7 4 SH9-SH11 duo-trio 83.3 5 SH9-SH14 duo-trio 83.3 6 SH9-SH 10 duo-trio 54.4 7 SH9-SH 13 duo-trio 36.4 Chasm RESULTS & DISCUSSION TOTAL SOLIDS & FAT DETERMINATION The desired fat and total solids content for the sherbet mixes were 1.5% and 34.2% and the actual values (Table 14) were similar, confirming that mix ingredients were added at the desired quantities and were pr0perly pre- pared. The percent fat and total milk derived solids of all sherbets meet the Table 14. Total solids and fat content of acid whey, WPC, sherbet mixes and sherbets. Saran]: Acid whey WPC Sherbet mix A Sherbet mix B Sherbet mix C Sherbet mix D Sherbet A Sherbet B Sherbet C Sherbet D W (%) * Eatmntent (%)" 6.781004% 5.531009% 34.291005% 34.6910.20% 34.2610.11% 34.641019% 34.0010.03% 34.631008% 34.4510.18% 34.541031% 0.031001% 0.111001% 1.521005% 1.571004% 1.481004% 1.511003% * Average of 6 replications for acid whey and WPC, 5 for sherbet mixes, 3 for sherbets. ** Average of 5 replications. 49 50 standards of identity (1%<%milkfat<2% and 2% 0.97) for all sherbet mixes and all replications (Table 17). The linear model (y = a+bx) gave slightly better r2 and was easier to work with. Therefore, this model was used for the data anal- ysis. Both the Newtonian and the Bingham plastic models are linear models, but for the Newtonian model there is no yield stress (00). Yield stress is the stress required to initiate flow, when applied to a fluid. In this case, the con- stant a played the role of the yield stress, and it was a very small number for all sherbet mixes (average values: 0.195, 0.097, 0.094, -0171 for sherbet mixes 56 A, B, C and D, respectively). Additionally, the negative yield stress calculated for sherbet mix D has no physical meaning. Practically a could be set at 0, and then equation (3) when so = 0, became equation (2) which described the New- tonian fluids, and up] became the viscosity 11 of the fluid. Constant b for the linear model in Table 17 describes the viscosity 11 of the sherbet mixes. This constant was statistically tested, in order to check if there was significant dif- ference among the average viscosity of each sherbet mix. A complete random- ized design and one way analysis of variance (AN OVA) was used. The AN OVA table for the viscosity of the sherbet mixes is given in Table H.1 in Appendix H [Larmond (1987), O’Mahony (1986)]. The viscosity values used for this analysis were in cP (where 1 Pa s = 1000 cP). Significant differences among the viscosities of the sherbet mixes at p<0.05 were found. The least significant difi'erence test, known as Tukey’s test, was used to determine which of the means for the viscosity of the sherbet mixes were significantly different. The mean values of the viscosities [for sherbet mixes A, B, C and D are given in Table 18. Larmond (1987) and O’Mahony (1986) describe the procedure followed in Tukey’s test. Table 18. Tukey’s test results for the viscosities of sherbet mixes at 40°F (4.44°C). Sherbet mix B A C D Average Viscosity“ 33.20a“ 28.97ab 26.90b 26.50b "' Average viscosity is the mean of three replications. ** Different letters next to means indicate significant difference at p<0.05. 57 At 5% level of significance the sherbet mix B was found more viscous than the sherbet mixes C and D, but no significant difi'erence was found among the rest of the samples. The viscosity of an ice cream or generally a frozen des- sert mix depends on its composition, especially the fat and the stabilizer/emul- sifier content. Guy (1978) found that the viscosity of ice cream mix containing whey solids from hydrolyzed sweet whey was decreased when whey solids replaced more than 50% of the MSNF. Patel and Harper (1977) found that ice cream mixes containing acid whey concentrated to 20% total solids by RO replacing 10-15% of the MSNF had higher viscosity than the control. The data for the present study showed a decrease in the viscosity from sherbet mix A to sherbet mix C and D, but the only significant difference was that between the viscosities of sherbet B with C and D. The higher viscosity of sherbet B may be attributed to its slightly higher fat content (1.58% versus 1.52, 1.48, 1.51% of sherbet mix A, C, D respectively) or to interaction of proteins at different levels of substitution of whey solids for MSNF. The sulfllydryl group of B—lactoglobu- lin has been implicated in the formation of complexes with x-casein upon heat- ing, in model systems. fl-lactoglobulin may penetrate the casein micelle, thereby inducing the formation of internal disulfide bonded complexes with x- casein upon heating [Farrell and Douglas (1983)]. Concentration dependency of the formation of these complexes may account for the peak in the viscosity of . sherbet mix B. 58 MELTING RESISTANCE OF SHERBETS The melting resistance of the sherbets at 38°C (100.4°F) is shown in Figure 3 as the volume of melted sherbet versus time. The data suggested that sherbet A melted faster than the other sherbets throughout the duration of the experiment (three hours). The graphs for sherbets C and D seemed to be very similar, at least for the first 150 min of the experiment, whereas sherbet B seemed to melt a little more than sherbets C and D after two hours. The over- all means of the volume of melted sherbet for all replications and all times for each sherbet were used to determine if there was significant difference in the melting resistance among the sherbets. The Tukey’s test (Table 19) was then used to determine which sherbets had significantly difi'erent melting resis- tance. The experimental model used for the statistical analysis was a com- pletely randomized design for factor A (different sherbets) with a split plot on factor B (time) [Petersen (1985)]. The AN OVA table for the melting resistance of the sherbets is shown in Appendix H (Table H.2). Table 19. Tukey’s test results for the melting resistance of sherbets at 38°C (100.4°C). Sherbet Average A B D C volume“ 97.946a" 85.929b . 80714b 80643b * Average volume is the mean of volumes of melted sherbet for all times and replications. *’“ Different letters next to means indicate significant difference at p<0.05. 180 150 _ 120 ml 90 Volume, 60 30 59 —o— Sherbet A --D— Sherbet B +Sherbet C +Sherbct D rrrrlfifi1 I 100 150 Time, min Figure 3. Melting resistance of sherbets at 38°C (100.4°F). 60 The results (Table H2 in Appendix H)) indicated that the melting resistance of the sherbets differ significantly at 0.1% level of significance. The results from the Tukey’ s test indicated that the melting resistance of sherbet A was significantly lower than that of sherbet B, C and D at p<0.05. No signifi- cant difl'erences in melting resistance among sherbets B, C and D were found. The resistance of sherbets to melting, therefore, was increased by the substi- tution of WPC for MSNF. The level of substitution was not so important, at least for replacement values more than 25%. The data suggested that the whey solids decreased the freezing point of the mix less than the MSNF decreased the freezing point of the control sherbet mix. So, even though all sherbets were held at the same temperature, those containing whey melted at a higher temperature. The higher level of lactose in the control would seem to explain this phenomenon. The whey particles may also act by binding the water particles, so again the melting of the sherbet could be delayed. The water holding capacity of the whey particles has also been reported in the lit- erature by Mathur & Shahani (1979). Increase of the melting resistance due to addition of acid WPC (obtained by reverse osmosis) in an ice cream mix has also been reported in the literature by Pater and Harper (1977). Huse et al. also found that ice cream containing WPC from sweet whey had slightly better resistance to heat shock than ice cream containing no whey. Rosenberger and Nielsen (1955) have reported increased melting resistance for ice creams con- taining spray dried whey powder in comparison to control ice cream containing no whey. A result of an increase of melting resistance would be the reduction of texture changes due to heat shocks that the final products might be subjected to during handling. This increase, however, should be within limits, so that the sherbet melts nicely in the mouth. 61 SENSORY EVALUATION OF SHERBETS The raw data of this experiment were analyzed statistically, in order to determine if there were differences among the acceptability of the sherbets and their iciness, creaminess and sweetness characteristics. For the accept- ability tests, the randomized complete block design was used and two way AN OVA was performed [Larmond (1987), O’Mahony (1986)]. The AN OVA tables for all the sensory tests are given in Appendix H. Flavor acceptability of sherbets Difference in the flavor acceptability of the sherbets was found at 0. 1% level of significance (Table H.3 in Appendix H). Sherbet B received the highest score, but it was not different from sherbet A (control), as the Tukey's test showed (Table 20). Sherbets A and B were found more acceptable in flavor than sherbet D, whereas sherbets A-C were not different. It should be men- tioned, however, that all scores were higher than 5, which in the 9-point hedonic scale means “neither like nor dislike” (Table C. 1). The highest score (7 .02) in the same scale means “like moderately”, whereas the lowest (5.69) is between “like slightly” and “neither like nor dislike”. To be rated acceptable and eventually marketed a product should receive a score of at least 7.0 on the 9—point hedonic scale [Ott (1992)]. Therefore, sherbet B seemed to be a prom- ising product for the sherbet market. Some of the most characteristic panelists’ comments follow. 0 For sherbet A: tasted great, was very creamy, needed more orange flavor, too sweet, could be fi'uitier, had good mouthfeel. 62 Table 20. Tukey’s test results for the flavor acceptability of sherbets after 8 days of storage. Sherbet B A C D Flavor score“ 7.02a** 6.56ac 6.15bc 5.69b * Flavor score is the mean of the scores received by 48 panelists. ** Different letters next to scores indicate significant difference at p<0.05. 0 For sherbet B: had good flavor, was very creamy, tangier than sherbet A, had more orange flavor than sherbet A but still needed more, sweet, did not leave bad aftertaste. 0 For sherbet C: had a weird aftertaste, tasted like cream cheese, was too creamy, good tangy flavor, good mouthfeel, flavor was a little off. 0 For sherbet D: tasted like cream cheese instead of orange sherbet, was creamy, had an unusual aftertaste, the sweetness is just great, very strong tangy flavor, more refreshing, had funny flavor, was the worst of all. All comments are presented in Appendix F. Although all sherbets received favorable and unfavorable comments for their flavor, sherbets C and D received more unfavorable comments. Panelists found that these sherbets had a unusual, funny, unpleasant flavor or aftertaste. Panelists mentioned that these sherbets tasted like Cream cheese, or orangy Cottage cheese and some others found their flavor terrible. There were, however, panelists who liked these products very much, because of their tart, fruity and not too sweet flavor. Table 21 shows the percentage of responses for the flavor of each sher- 63 bet in each category in the 9-point hedonic scale. Fifty six and twenty five hun- dredth percent of the panelists liked sherbet A moderately to extremely, whereas sherbets B, C and D were liked modemme to extremely by 75, 54. 17 and 47.92% of the panelists, respectively. This comparison shows that even sherbets C and D were considered acceptable by half of the panelists and could potentially be marketed. Table 21. Percentage number of responses in each category of the 9- point hedonic scale for the flavor of each sherbet. Scale (% responses) (%responses) (%responses) (%responses) Like extremely . 6.25 20.83 8.33 4.17 Like very much 25 25 22.92 20.83 Like moderately 25 29.17 22.92 22.92 Like slightly 18.75 10.42 14.58 8.33 Neither like nor dislike 14.6 0 2.08 8.33 Dislike slightly 8.3 8.33 18.75 18.75 Dislike moderately 2.1 4.17 4.17 8.33 Diser very much 0 0 4.17 4.17 Dislike extremely 0 2.08 2.08 4.17 Texture acceptability of sherbets Difi‘erences in the texture acceptability of the sherbets at 0 1% level of significance were found (Table H4 in Appendix H). Table 22 summarizes the results of this test. - 64 Table 22. Tukey’s test results for the texture acceptability of sherbets after 8 days of storage. Sherbet Texture B A D C score“ 7.429“ 7.33s 6.31b 6.13b * Texture score is the mean of the scores received by 48 panelists. ** Different letters next to means indicate significant difference at p<0001. In texture, sherbets A and B were found significantly more acceptable . than sherbets C and D. The average scores for sherbets A and B were close to “like moderately”, whereas the average scores for sherbets C, D were close to “like slightly”. Sensory analysts consider scores such those received for sher- bets A and B good enough for marketing of a new product [Ott (1992)]. From the panelists’ comments and the researcher’s observation, it seemed that the whey solids imparted to the sherbet a crumbly body. Sherbets C and D tore apart when scooped. Some of the most common comments obtained by the sen- sory panel follow. All verbatim are given in Appendix F. ' For sherbets A, B: were very smooth and creamy, had pleasant textu- re, were soft, spooned well, had more like ice cream texture, melted nicely in mouth. sFor sherbet C: not as smooth as it should be, hard, icy consistency, texture good for sherbet, spoon went in hard, crumbly texture, did not melt in mouth smoothly. 0 For sherbet D: Not as smooth texture, icy, gritty, felt rough to the tongue and roof of mouth, hard, more frozen, broke apart, did not melt in mouth as easy as the others. 65 The observations concerning the crumbly body of sherbets C and D can be explained by the difference in protein composition of the MSNF and WPC. Milk solids non-fat contain approximately 27% casein and 8% lactalbumin, whereas WPC contain no casein and 13% lactalbumin. Casein gives body or substance to dairy products, such as ice creams or cheese, whereas lactalbumin is not able to contribute the same tactual properties and, in fact, after being heated in processing (for frozen desserts in pasteurization) imparts to the food preparation a short or crumbly body [Carter et al. (1982)]. However, at 25% level of substitution of WPC for MSNF the texture acceptance of the sherbet did not differ from control. The harder texture of sherbets C and D observed by many panelists can be attributed to possible higher freezing point of these products. Higher freezing point results in harder products, which are not easily scooped. Good texture, however, is described differently by different people, so there were panelists in this study, who found the texture of sherbets C and D more accept- able, because it was icier, less smooth and the products did not melt very fast, whereas others liked sherbets A and B more, because they were smoother. Table 23 shows the percentage of the number of responses for each category in the 9-point hedonic scale for texture. The texture of sherbets A and B were liked moderately to extremely by 81.25 and 83.33% of the panelists, respec- tively, whereas the corresponding percentages for sherbets C and D were 47.91 and 54.16, respectively. The fact that about 50% of the panelists liked the fla- vor and the texture of sherbets C and D moderately to extremely may indicate two target markets with two different products. 66 Table 23. Percentage number of responses in each category of the 9- point hedonic scale for the texture of each sherbet. Wain W W Sham W scale (% responses) (%responses) (%responses) (%responses) Like extremely 20.83 10.42 6.25 8.33 Like very much 37.5 45.83 20.83 18.75 Like moderately 22.92 27.08 20.83 27.08 Like slightly 10.42 10.42 18.75 22.92 Neither like nor dislike 0 4.17 8.33 2.08 Dislike slightly 2.08 2.08 16.67 12.5 Diser moderately 4.17 0 6.25 4.17 Dislike very much 0 0 2.08 2.08 Diser extremely 2.08 0 0 2.08 Iciness, creaminess and sweetness comparison of sherbets The iciness, creaminess and sweetness panels consisted of 6, 5 and 5 trained panelists, respectively. In order to obtain more data and make the tests more reliable all multiple paired comparison tests were performed in triplicate. The data were treated as if they had been obtained by 18, 15 and 15 panelists for the iciness, creaminess and sweetness test, respectively. To make sure . that there were no significant sample-panelist interactions, the results were analyzed first by analysis of variance with interaction [O’Mahony (1986)]. This method checks whether the panelists are consistent with their answers on a specific pair of samples among the three replications. The anal- ysis of variance table for the AN OVA with interaction for the iciness compari- son test is given in Appendix H (Table H.5). The analysis showed that the interaction sample-panelist (axA) was not significant at 5% level of signifi- cance. This was another proof of the trained panel’s ability to distinguish between difl‘erences in iciness. The regular analysis for the Schefi'e test could 67 be performed. An average value for each sherbet was calculated in this anal- ysis. The values were relative and their sum for all sherbets must be 0 The exact procedure for the data analysis are given by Larmond (1987). The Tukey’s test table is given in the following: Table 24. Tukey’s test results for the iciness of sherbets after 9 days of storage. Sherbet Iciness D C A B score“ 0.720a** 06403 - 0399b - 0961b * Iciness scores are reported as main effects of treatments [Larmond ( 1987)]. ** Different letters next to means indicate significant difference at p<0.01. Significant differences in iciness among the sherbets were found at 0.1% level of significance (Table H.6). The order efi‘ect, however, was found not significant. The panelists found sherbets D and C icier than sherbets A and B. The higher iciness of sherbets D and C was easily observed even when the samples were spooned or scooped. The fact that sherbets D and C were icier does not mean that they were bad products. In fact some panelists found the least icy samples smoother than a sherbet should be. Characteristic panelists’ comments are given below. These comments are in agreement with the com- ments obtained by the untrained panelists who judged the sherbets for texture acceptability. 0 Sherbets A, B are very fine products. O Sherbets C, D have slightly detectable iciness, but are both fine and smooth and are very good products. 68 'Sherbet D is much icier than B, but it is not an icy sherbet, it is quite fine and smooth. “Even the appearance and taste of D is icier. °Sherbets C and D feel colder. oI prefer the texture of C better (compared to A and B), even though it is icier. H i ’ i i 0 Sherbet B is too smooth. The addition of WPC, therefore, in the sherbets increased their iciness. The iciness is influenced by the STEM content, the mixing conditions during freezing of the mixes and the temperature conditions during storage. In this study, the above factors were the same for all sherbets. The different composi- tion of the WPC and MSNF might have led to this difference in iciness. The protein composition of the WPC is probably responsible for the crumbly body of the sherbets with 50 and 75% substitution of WPC for MSNF. The same factor might be a possible cause also for the difi'erence in iciness. Sherbet B, however, was found not significantly different than the control sherbet A in iciness. So the substitution of WPC for MSNF at a 25% level did not influence the iciness of the final product. The results for the sweetness test were also analyzed by the ANOVA with interaction and no significant interaction axA was found. The order efi'ect was found not significant, whereas the main effect was significant at p<0001 (Table H.7), so the sweetness of the sherbets was different at this level of sig- nificance. The Tukey’s test (Table 25) showed that sherbet A was the sweetest, although not significantly different than sherbet B. Sherbet A was sweeter than. sherbets C and D and sherbet B was sweeter than sherbet D, whereas sherbets B-C, C-D were not different in sweetness. 69 Table 25. Tukey’s test results for the sweetness of sherbets. Sherbet Sweetness A B C D score“ 07928“ 0170ac - 0411bc - 0.551b * Sweetness scores are reported as main effects of treatments [Larmond (1987)]. ** Different letters next to means indicate significant difference at p<0.05. The difference in sweetness can be attributed to the increase of the acidity of the sherbets as the level of the replacement of MSNF by WPC increases. The tart flavor covered the sweetness of the sherbets, so that the control sherbet containing no whey was the sweetest, whereas those with the lower pH were the least sweet. Some of the panelists preferred the least sweet samples, because they found their tartness more acceptable for a sherbet prod- ~ uct. Some panelists found the sweetness of sherbet A and B more intense than it should be. Since sweetness is an important constituent of the flavor of a fro- zen dessert, it should have played an important role in the flavor acceptability of the sherbets. This was shown by the large number of the untrained panel- ists’ comments (Appendix F) related to the sweetness of the samples. Some of. the trained panelists’ comments follow. ' Sherbets C, D are both sour. ' Sherbet D has a sour taste, that may mask the sweetness. 0 Sherbet D was a good in sweetness product. 0 Sherbet A was extremely sweet. 0 Sherbet D has a more orangy, fi'uity flavor. . 70 The substitution of WPC from acid whey for MSNF resulted in less sweet and more tart products. The same sherbets were also found fi'uitier, which suggested that the acid enhanced the orange flavor. Another flavor and texture constituent of frozen desserts is creami- ness. Analysis of the raw data by AN OVA with interaction showed that there were no panelist-sample interactions. The presentation order of the samples was found not significant, whereas the sherbets were found different in cream- iness (p<0.001) (Table H.8). The Tukey’s test (Table 26) showed that sherbets A and B were creamier than sherbets C and D (p<0.01). The WPC addition to the sherbet mixes seems to influence the creaminess intensity of the sherbets at levels of substitution higher than 25%. The panelists’ verbatim showed that sherbets A and B were creamier than the average commercial sherbet. The WPC may have covered the creamy feeling of the sherbets by making them more refreshing, due to their higher acidity, lower sweetness and higher ici- ness. The products with high WPC content, therefore, seemed less rich. The comparison of the creaminess data with the rheological data (Tables 26 and 18) suggested that sherbet B was creamier and more viscous than sherbets C and D, whereas sherbet A was creamier but not more viscous Table 26. Tukey’s test results for the creaminess of sherbets. Sherbet Creaminess A B D C score“ _ 0.788a 0567a -0554b -0801b "' Creaminess scores are reported as main effects of treatments [Larmond (1987)]. ** Different letters next to means indicate significant difference at p<0.01. 71 than sherbets C and D. A creamier sherbet had been expected fi'om a more viscous sherbet mix. The data, however, are not enough to support this assumption. Further research is necessary for the investigation of the rela- tionship between creaminess and viscosity. HEAT SHOCK STABILITY TESTS Differences among the sherbets in flavor acceptance (p<0.01), texture acceptance (p<0.001) and iciness (p<0.001) were found (Table H.9, H. 10 and H11). The Tukey’s test (Table 27) showed that sherbet C was more acceptable in flavor than sherbet D (p<0.01), but no other differences among the samples were found. All scores were higher than 5 ("neither like nor dislike") and lower than 7 ("like moderately") and no flavor deterioration was reported by the pan- elists. The panelists’ comments for the flavor of the heat shocked sherbets were similar to those obtained in the first flavor acceptance test. The texture accept— ability of the heat shocked sherbet D was lower than that of the other sherbets. The average texture acceptance scores (Table 27) indicated that, even after the heat shock treatment, sherbets A and B had acceptable texture (scores > 7). Therefore, the heat shock treatment did not measureably influence their tex- ture. Sherbet C, however, was found not different from sherbets A and B in both flavor and texture acceptance tests. This is probably due to simultaneous changes in the flavor and texture of all sherbets. For the iciness of the sherbets, AN OVA with interaction showed that there was no panelist-sherbet interaction. The iciness mean scores indicated that all sherbets, except sherbet B, were not different in iciness. Sherbet B was found the least icy. 72 Table 27. Tukey’s test results for the flavor and texture acceptance and the iciness of sherbets after a heat shock treatment. Sherbet A B C D Flavor score" 6.29ab*** 6.31ab 6.71a 5.50b Texture score“ 7.15a*** 7.23a 6.71a 5.56b Iciness score“ 0.028a*** -0736b 0167a 0542a * Flavor and texture scores are the means of scores received by 48 panelists. ** Iciness scores are reported as main effects of treatments [Lamond ( 1987)]. *** Different letters next to means indicate significant difference at p<0.01. STORAGE STABILITY TESTS The flavor and texture acceptance scores and the iciness scores for the sherbets after the storage period are given in Table 28, whereas the corre- sponding AN OVA tables are shown in Appendix H (Tables H.12-H.15). After 31 days of storage (Table 28) sherbet A was found more acceptable in flavor than sherbet C only (p<0.05), whereas all the other sherbets were found not dif- ferent. The scores for all sherbets were lower than 7 and higher than 5 (between “like moderately" and “neither like nor dislike”). After 122 days of storage, sherbets A,IB and C did not differ in flavor acceptability and only sher- bet D was found less acceptable in flavor (p<0.05). The texture scores of sherbets A and B were above 7 ("like moder- ately"), which indicated that the 31 days of storage did not change the degree of liking of their texture. Sherbet C, however, was found not different in tex- ture acceptability fi-om sherbets A and B. 73 Table 28. Tukey’s test results for the flavor acceptance, texture acceptance and iciness of sherbets after storage. Sherbet A B C D Flavor score“ after storage for 31 days 6.65a**** 6.27ab 5.73b 5.85ab 122 days 6.85a**** 7 .42a 6.88s 5.88b Texture score" after storage for 31 days 7.27a**** 7.31a 6.73a 5.85b Iciness score*** I after storage for 32 days -0547c**** -0656c 0.848s 0355b * Flavor score is the mean of the scores received by 48 panelists. ** Texture score is the mean of the scores received by 48 panelists. "*Iciness scores are reported as main effect of treatments [Larmond ( 1987)] **** Different letters next to means indicate significant difference at p<0.05. 74 For the iciness test performed after 32 days of storage no panelist-sam- ple interaction was found. After this storage period sherbets C and D were still the iciest. The data suggested that sherbets A and B were not different in iciness, so if there was any change, it was probably of the same intensity for both samples. This was in agreement with the texture acceptance results, where sherbets A and B were not different. Sherbet C, however, was found the iciest in this test, although its texture was found to be at the same degree acceptable as that of sherbets A and B. The texture acceptability is very closely related to the iciness of the sherbets, however it is not related only to this attribute. Some of the panel- ists’comments showed that other attributes, such as creaminess, hardness and resistance to melting are also very important for their decision. The influence of these other attributes may have resulted in this disagreement between the texture acceptance and the iciness results for sherbet C. The panelists’ com- ments for the iciness of the sherbets after 9 and 32 days of storage were simi- lar, which indicated that the iciness changes were probably small. COMPARISONS The average scores for all the flavor acceptance tests (after 8, 31, 122 days of storage and after the heat shock treatment) and all the texture acceptance tests (after 8, and 31 days of storage and after the heat shock treatment) are shown schematically in Figures 4 and 5. Comparison among the data of Figure 4 shows that even after 122 days of storage at 0°F (-17.8°C) and after a heat shock treatment the flavor acceptability of the sherbets was always between “like moderately” and 75 “neither like nor dislike”. However, in all cases, sherbet D had scores lower than 6 (" like slightly"). The flavor of a sherbet depends on many factors, such as the sweet- ness, tartness, richness, e.t.c. Although the sherbets difl’ered in some of these flavor constituents, their flavor acceptability was not that different. This could be attributed to the fact that different people have different perception about what good flavor is. Therefore, they scored the sherbets based on this percep- tion. The important point from this study of the flavor acceptability of the sher- bets was that no product had a noticeable spoiled flavor due to the heat shock Flavor score NMAUICAQOOO p—s Sherbet Figure 4. Flavor acceptance of sherbets after 8 ( i ), 31 (E3 ), 122 (um ) days of storage and after the heat shock treatment ( ). (1 I: dislike extremely, 9 = like extremely on the 9-point hedonic scale). 76 ' treatment or the storage. Only one panelist found the flavor of sherbet A "not as fresh", after storage for 122 days. The "funny" or "unusual" taste or after- taste of sherbets C and D reported by many panelists was not developed during the heat shock treatment or the storage, but it was a characteristic that these sherbets also had when they were first evaluated. Apparently, these undesir- able characteristics had their origin in the whey itself. Texture score NUJALIIO‘QOOO H Sherbet Figure 5. Texture acceptance of sherbets after 8 ( 52: ) and 31 ( ) days of storage and after the heat shock treatment ( ). (l = "dislike extremely", 9 = "like extremely" on the 9-point hedonic scale). 77 In all cases sherbet B had slightly higher texture scores than A (Figure 5), although not significantly different. Sherbet D was always less acceptable in texture acceptability than sherbets A and B. The importance in this compar- ison was that sherbet C, after storage for 31 days or after the heat shock treat- ment had texture acceptability not difi'erent than sherbets A and B (Table 28 and 27). This could be attributed to a simultaneous change of the texture of all sherbets, which was more intense for sherbets A and B. This explanation, for the texture scores after the heat shock treatment, is in agreement with the melting resistance results, where sherbet A was found to melt significantly faster than the rest of the sherbets. So, during the heat shock treatment of the sherbets, sherbet A melted the most and, therefore, it was subjected to more intense changes in texture in comparison to the other sherbets. This probably resulted in a product at the same degree acceptable in texture as sherbet C. It should be noted, however, that the texture acceptabilities of sherbets A and B were always higher than 7 in the 9-point hedonic scale, which indicated that even after the heat shock and 31 days of storage these products were consid- ered marketable. Some untrained panelists, however, reported that sherbet A had many ice crystals after the heat shock treatment. Therefore the verba- tim from the texture tests after the heat shock treatment are presented in Appendix F, separate from the other comments. Table 29 summarizes all the iciness results. After the heat shock sher- bet A became icier, as the data suggested, (some of the trained panelists’ com- ments were also showing that), and it was not different in iciness from sherbets C and D. Since its resistance to melting was found the lowest, it melted the most, and therefore it had the highest textural change. After the sherbet was put back again at 0°F, it probably developed larger ice crystals than those it originally had, which made it comparable to the iciest sherbets, C and D. 78 Table 29. Tukey’s test results for the iciness of sherbets after 9 and 32 days of storage and after a heat shock treatment. . Sherbet Iciness score" D C A B after storage for 9 days 0720a" 0640a -0399b -0961b storage for 32 days 0.355b** 0848a -0.547c -0656c heat shock 0542a” 0166a 0028a -0736b * Iciness scores are reported as main effects of treatments [Larmond (1987)]. ** Different letters next to means indicate significant difference at p<0.05. FURTHER DISCUSSION ON THE PANELISTS’ COMMENTS The most common comment for the flavor of sherbets A and B was that they were very sweet, very creamy and did not have enough orange flavor. They were characterized as bland by many panelists. In contradiction, sher- bets C and D were found better in sweetness, acid and orange flavor intensity, but there were many comments about an undesirable aftertaste, a Cream cheese flavor and, in some cases, a bitter taste. Most of the panelists were able to detect that there was something unusual added to sherbets C and D. The bitter flavor can be attributed to the amino acids, the peptides and the higher content of calcium salts contained in WPC [McGugan et al. (1979)]. A dry feel- ing after tasting sherbets C and D was also reported, which is probably due to the higher water holding capacity of the WPC. The above comments show that sherbets A and B could be further improved in flavor, by adding more citric acid and probably orange sherbet base in their mixes before freezing and/or by reducing the sucrose content. 79 This would increase the tart and orangy flavor, which was very desirable, and decrease the sweetness intensity. Addition of citric acid would also increase the titratable acidity of sherbet A, which was lower than the standard value of 0.35% for sherbets [CFR (1990c)]. For sherbets C and D, however, the Cottage cheese whey flavor was very noticeable, and probably only the addition of more orange sherbet base (to cover the whey flavor) could make these products more acceptable in flavor. The amount of citric acid added to sherbet D could be slightly decreased, since many panelists found this sherbet too tart. A different flavoring system could also probably better cover the whey flavor of sherbets C and D and improve their flavor acceptability. Most of the panelists found sherbets A and B too soft, and some com- mented on the very fast melting of sherbet A. These two sherbets would prob- ably be very good, if served in a dish, since it was reported that they were easily spooned and scooped. Sherbets B, C and D could be also served on a cone, since they do not melt fast. It should be taken into account, however, that sherbets C and D tore apart easily, which is not an acceptable characteristic for a fi'ozen dessert. The panelists’ comments suggested that sherbets A and B were softer than sherbets C and D. As previously discussed, the hardness of a frozen des- sert is related to its freezing point. Since all sherbets were kept at the same temperature, the comments about hardness suggested that the fi'eezing point of the sherbets increased fiom sherbet A to sherbet D. The fi'eezing point is related to the soluble solids content of the mixes. The higher lactose and salt content of the MSNF probably resulted in a higher fieezing point depression than the WPC, so the sherbets containing more WPC were harder. Chapter 5 CONCLUSIONS AND RECOMMENDATIONS This study suggested that the substitution of direct-set Cottage cheese whey ultrafiltration retentates for 25% of the MSNF in orange sherbet resulted in a product (sherbet B) of similar flavor and texture acceptability, iciness, creaminess and sweetness to the control (sherbet A) containing no whey solids. This product received flavor and texture acceptability scores good enough for consideration of marketability. The sherbets with 50 and 75% substitution of whey solids for MSNF (sherbets C and D) were less acceptable in flavor and texture and icier in comparison to the control sherbet and the sherbet with 25% level of substitution. But they had a tart and less sweet flavor and an icier tex- ture which were desirable characteristics for 50% of the panelists. The fact that about 50% of the panelists liked the flavor and/or texture of these prod- ucts moderately to extremely indicates that there are probably two potential markets, one for sherbets A and B and one for sherbets C and D. All sherbet mixes exhibited Newtonian fluid behavior at low shear rates. Sherbet mix B was found more viscous than sherbet mix C and D, but no other differences in viscosities were found. The use of acid whey in the for- mulation of the sherbets increased their melting resistance and protein nitro- gen content. Increase in protein nitrogen content, and therefore protein content, would be of interest from the marketing point of view. 80 81 N o flavor deterioration attributable to the heat shock treatment or the storage for 31 and 122 days was reported by the panelists and sherbets A and B received good texture scores (greater than 7 in the 9-point hedonic scale) even after the heat shock treatment or the storage for 31 days. These results indi- cated that sherbets A and B were quite stable products. The potential marketing of all three sherbets containing whey, or at least of sherbets B and C should be further investigated. Changes of the com- position of the sherbets, such as decrease of the amount of sweeteners in sher- bets A and B or increase of the amount of citric acid added to them and decrease of the amount of citric added to sherbet D would probably result in more accept- able products with higher market potentials. The use of a difl'erent flavoring system or modification of the system used could also result in more acceptable products. The same research could be conducted by keeping the amount of acid constant in all sherbets, in order to eliminate the influence of the different sweetness and acidity on the product flavor. Another area that needs to be further investigated is the relationship between the viscosities of the sherbet mixes and the creaminess of the final sherbet products. More data are necessary for such a relationship. Addition- ally, the presence of protein interactions and their influence on the viscosity of the sherbet mixes could be studied. The fieezing points of the mixes should be determined and correlated to the melting resistance and iciness results. There are equations that predict the freezingpoint of every mix, when its exact composition is known and could be used to make comparisons with the experimentally determined values. Finally, the economics of the whole procedure should be investigated, taking into consideration the capital and operational costs of the ultrafiltration process itself, the cost of the membrane replacement, the cost of the milk solids 82 non-fat in comparison to acid whey and the cheese manufacturers’ savings resulting from the minimal treatment of whey before it is used in a sherbet for- mulation. APPENDICES APPENDIX A DRY WHEY ANALYTICAL STUDY 83 £54 c~.c «5.: n~.o cw.“ n~.c oo.o =~.o n=.: cm.c mm.— n5.c on.o em.c an.c =~._ n~.o 5~.c a... o:_u>dv oquou; 2.1.3.33: 1_u< 0.5.3.33 u:._e> o:_uom a:_5m>u o=_::a.:h 9...... _ < 1_u< oua.=:z< o=_:u_=_>:a:; ..e._..3.;..._. e:_eccab a.._:_”_.< u:_u:e; e:_=_5m_: o:_u=:._om_ u:_m>; Am5c5wv m=_u< ::_&€ me.o __.o oc.o ~c.o n.— c.o =.~ o.o 5.— «a. an:— ec~_ c_o_ ¢- Assay u¢=uau< nasav a:.5==u >513 an: manna no. no; asa~a>5 .255V s=.=o.em .5. “can o:..o;5 Assn. ausuuoz .c.o Amsv 5_u< u.~om .25e5 555. 5.. x55. 5.55.: Assay azaaon a.mn “rosy :_.o_= AEmzv common 5... Anew aqu< u_:5;uo.=5m .521 55.. 5.5 .5555 N.235238 .551 55.5 5.5 x5e. 55\cc.aon...m Ansv e=.oo:5nz ..~ x551 ~5.c.sc..on.5 x5e. 55.555555 5.5 Anew .5\e=.sa.;e .551 55.555 55.5 .5:. m 1555 55.5555555 5.. .525 u .5sv 25.5.55 an. A.=..V < 5.5.qmmm mmmammqm .555: 5:4. 3...: a... 25 .5003... «c man—canoe Eve cum—nu one 3.52:8 .5855 cashmere. .—.< 03¢? £14 c~.o «5.: n~.e 55.. n..c 55.c =~.o «5.: 55.: mn._ n5.o 95.: en.o an.c =~._ n~.o 5..c a... e:.u>.3 o... «9...— 2.1.3.3.... 1.“: 35.3.3”. e:..e> e:_.em 3:..mxu 5:.235515 e:..:._< 1—u< u—..=:a< 9......— .. _ ass...- :=:;o.;>.+ u:.eo..b c...:.._.< a... 5...... e=_=..m_= a:.e=:_am_ e:_a>; .mse.m. m=_5< e:_&¢ .513 .5: 52552 no. uue ne=.mpw mo.o Ages. udzoau< ._.5 .255. 25.2555 oc.o Aaazv 2:.zuaem _c_ .12. 0:..o:u 55.5 .255. .555552 .5.5 .52. 5.5< 5..55 5.. .255. 555. n.. .55. 5.55.: 5.5 .255. 55.55. 5... .552. 5..5.5 =.~ .551. 599509 e... Ame. suu< u.:a:.o.eum 5.5 .52. 555. 5.. .552. N52.25.5555 .._ Awe. u:.~ c.o Ace. ca\u:_xce..xm ~o. Ana. szqoezaa: _.~ .15. ~a~c~>a—.cn—¢ 5.5. .52. 25.555555 5.5 .52. .2.55.25.55 55.. .52. 55.555 55.5 .55. 5 a_o. Ame. 5:56:15611 a.. .92. u 5.. .52. 25.5.55 .n. ..5... < n.2aoflflz. mw_Ecu_> .55.... .59.. 52.55 .5... 2.... .5835 me 3:35:00 30.. cum—5. and 3.55.: .5353 095.595. .n.< 03.3.. 85 o~.o ...o _~.c -.~ n5.o .n.o 5~.o 55.5 55.5 a... .3... 5~.o .n.o .m.c 5~.. .n.o 55.o m... a:.u..c 9:..351 o:_:o.;.5: aqu< u.§:.:_o o:..a> 9:..sm a:..a.u a:.:oo.;h ::_::.< 5.55 5...555< e=.:a_=.>:e;a ..=._._c..;..._. s:.mocxb o:_:.c.< o:.u:o. e:.c_ue_= 5:.5:9_cu— e:_m>; Amasnw. m1.u< ozmafi am.o e..c ma.o a.— c.a 5.5 sew new. NNQ. c.m_ a.- .2;;. 5.:555< 55:1. 5:.Evcc Aaaz. 52.:9—am A511. .5:555: A...._._. cam.— Aszz. 0:.15— A519. muzzau .25. :35. A25. 5:.N .25. 52.55215: .25. 92.55555. .25. E:_1om Ana. 5:50:1501m Ame. Esau—no M—sauzam .51; .51 525.5 co. .5; ao:.a>c m.o mo.o n.o .o. .52. 55..555 .52..5.5< 5..55 . Ame. c.5n.= Anus. :.5c—n .52. 5.5< 5.55;.5.555 Anus. ~_=\=.E=.aaou has. cm\o:.xon.5xm Awe. ~=\c_>a..on.¢ has. _=\5:.35.:h .52. 5 .52. 5 ..5... < m: 153...... .55.... 595. 55...: .3. 5...... .30.. no 3:05:60 30.. 05.5... was .555:— .5533 ougo>< d4 035B APPENDIX B PRODUCTION AND UTILIZATION OF WHEY 86 87 Table B.l. Estimated U.S. fluid whey and whey solids production by type and resulting quantity of whey solids further pro- cessed [ADPI (1991)]. 1925 1_982 12.82 1.991 2 19882 Sweet-type Uhev Production 1 Cheese Production’ 6,674 5,025 5.209 5,344 5.572 Calculated rluid whey“ 42.066 45 . 255 to . 881 48.096 50,148 Calculated whey Solids’ 2,73a 2,950 3,057 3.126 3.259 Acid-tzpe Uhev Production Cottage Cheese Production’ 603 960 970 945 938 Calculated Fluid whey“ 3,613 5.760 5.820 5,670 5.62. Calculated "hey Solids“ 235 376 378 369 366 Total Whey Production (fluid basis): 52.611152521215121 M Total Whey Production (solids basis): 2.969 3.316 3.325 3.495 3,625 who! Solids Further Processed: A-Concentrated whey Solids“ 130 51 47 29 37 B‘Dry Uhey - Human food 725 812 890 866 940 - Animal Feed 173 l75 141 231 197 C-Hodified Dry Whey Products - Reduced Lactose Hhey 8 60 96‘ 90' 107' 122' Reduced Minerals Whey ‘28 - whey Protein Concentrate” 96 105 78 96 136 D-Uhey Solids in we: Blends 136 136 116 134 128 Z-Hhey Solids Utilized for Lactose“ 198 l97 216 242 258 Total "he: Solids Further Processed (A+8+C+D+B): 1.546 1.572 1.578 1.706 1.818 Total "he Solids further Processed as 1 of Total Uhev Production solids basis : 52.15 47 41 b6 12 48.81 50.2! III-II .— 1 \kiunmrfigur-rhrndnfinrnlb ' 2 Ihnfised. 3 lupflmflunufléhafisfiarBbanLliASS,LEHMA. 4 Iflhuanroducflon:spprurhnstdbrOlbfllniches-e[Inducad(sxeept(hntegel AppnndaumflyWIHVln>CouagechuuB|nnducul JuanagpflnlalSofldscnntsntofntnyzdifli. lumuagplhnalBonds(humantokallnnunnedlvhey:40In IDIUInotavaflabkh lflsdueedlaunoseluulinducedllfinsnehIVthycouinnedtoswuhidildosunsofindhddualyflentopsrefibn. .RspmnadaaIhwfianyIlflaumlaflflhnnnenflhmdlflhqythnnqfll1981. 0 AquanxhnauflwfiLCIbslvhqysofidsruflhudVllblaculniuodquL OOQOOI fl 88 no.oo_ o.nnn N~.o ~.m N_.c n.o N_.c n.o N~.o m.— N..: «.0 Nn.o n.~ N_.o 0.6 N~.n s._o um.o o.n N.. m.c Nm.o _.< no.c ~.o N~.c— n.-_ «0.. 0.5 N~.o n.n Nm._ n.__ Na..— n.mo N¢.c o.ne Rn.¢< ~.nnn Nu.m c.n< c.0— o.~cn eve—>eu I scouted “nook \. ona~-v.u< onah-uoosm .evcsod uo ecodudal cu sensuau elsuo>‘m .neuee uueuuvc— use uueuve sue; nevauocuMw no.oo_ ..~o~ o.~ ~.enh snoop an.c .mqmun mqm ~.~ n I u a Ova-Ian 5.33 a: uvadom >053 u_.o ~.o - ~.o ou-tucoueou cannot; sus: u_.o o.c - a.o gaseous hog: eo.ta N~.o ¢._ - a._ has: co..o - - - . >05: veaeuuceucou nooou noguo n_.o ..c - c.o unsung use: :. nesaom sag: NN.° .1— n o‘— GHIHuCOUCOU :‘Oubum hO—s . - - . . auauoum hes: teat: an.« o.~n - a.~n has: gusto u~.c 5.. c 5.. »e33 veg-cuceocou nooom you no.. o.~ - a.“ accede sus: cs ucqaom has: ne.o n.q - n.¢ canton-ucoo c.ouo.m use: u..o _._ - _._ uuaootm has: ao..a u_._n _.~n~ a.c n.on~ ass: eo_to no._ o.~ . 0." ans: neg-aucoucou evoou ocusm n_.o _.o - _.o mucosa sag: c. oas~om hog: - a - . eueuuceucou :«euoum avg: . - - . uostCm >033 veaua um.o _.¢ - _.e hog: so..o - - - . hog: veaeuuceucou n—VUUK Had—SOL "0.. _.~_ - _.~_ noses. soc: c. .u..om has: uo.o {.mh - e.n~ «autos-ucoo cogent. sus: N¢.a ~..a - ~.en uuaeo.a tog: eo_.a um.on _.¢- o.~ _._a~ 52.: gusto no.5 e.¢m - o.om .n« has: was-tocoucou nuns. ososuo\u..u\a...a azouuom deuOP enafl-v—u< onaH.uueam Nuance-u uexuez cza— Irnblu— cu: .53: 5:3 «8368 82 o5 9:: cc Engage .amcoou 1335 am 3035.:— hoaB and hog—B no flotsam—GD dd 039—. APPENDIX C QUESTIONNAIRES FOR THE SENSORY EVALUATION TESTS 'IVvo questionnaires for a flavor acceptance test are given here. The questionnaires for the texture acceptance test were similar. The first of the two questionnaires given here was presented first - third to the panelists, whereas the second was always the questionnaire given with the last sample to each panelist. For the comparison tests, one questionnaire for iciness is given here. The questionnaires for the other attributes tested were similar. For these questionnaires, the sample code numbers were hand written, because they were different for each panelist and for each session. Note 1: The scores corresponding to the 9-point hedonic scale divisions (for the acceptance tests) are written on the first questionnaire next to the scale. Note 2: The scores corresponding to the 7-point Schefi'e pair comparison scale (for the compari son tests) are written on the first questionnaire next to the scale. 89 90 Table C.1. Questionnaire (l) for the flavor acceptance test. Panelist #: Name: Date: Characteristic tested: Flavor acceptance I INSTRUCTIONS 0 You are given a coded sample of orange sherbet. Before tasting the sample, please rinse your mouth with some water. 0 Check how much you like or dislike the flavor of this sample. Code: 511 _like extremely (9) _like very much (8) _like moderately (7) _like slightly (6) _neither like nor diser (5) _diser slightly (4) _diser moderately (3) _diser very much (2) _dislike extremely (1) Comments: Afier finishing this portion of the panel, please show the “ready" card in your booth. Another sample will be served to you. 91 Table 0.2. Questionnaire (2) for the flavor acceptance test. Panelist #: Name: Date: Characteristic tested: Flavor acceptance INSTRUCTIONS 0 You are given a coded sample of orange sherbet. Before tasting the sample, please rinse your mouth with some water. 0 Check how much you like or dislike the flavor of this sample. Code: 511 _like extremely _like very much _like moderately _like slightly _neither like nor dislike _dislike slightly _diser moderately _dislike very much _diser extremely Comments: When you finish, show the “finished” card in your booth. W very much for your time and consideration. 92 Table 0.3. Questionnaire for the iciness comparison test. Panelist #: Name: Date: Characteristic tested: Iciness INSTRUCTIONS 0 You are given two coded samples of orange sherbet. Before tasting each sam- ple, please rinse your mouth with some water. 0 Taste the products in the following order: 575, 392. 0 Examine these two samples of orange sherbet for iciness. 0 Indicate the degree of difference in iciness between the two samples by checking one of the following statements. 575 is extremely icier than 392 __ (+3) 575 is much icier than 392 __ (+2) 575 is slightly icier than 392 __ (+1) no difference __ (0) 392 is slightly icier than 575 _ (-1) 392 is much icier than 57 5 __ (-2) 392 is extremely icier than 575 (-3) Comments: APPENDIX D HYPOTHESES FOR THE SENSORY EVALUATION TESTS WV” acceptability among sherbets A, B, C and D. Alternative: There is difference in flavor acceptability among sherbets A, B, C and D. W Null: There is no difference in texture acceptability among sherbets A, B, .C and D. Alternative: There is difference in texture acceptability among sherbets A, B, C and D. Null: There is no difference in iciness among sherbets A, B, C and D. Alternative: There is difference in iciness among sherbets A, B, C and D. 93 Null: There is no difi'erence in sweetness among sherbets A, B, C and D. Alternative: There is difference in sweetness among sherbets A, B, C and D. Null: There is no difference in creaminess among sherbets A, B, C and D. Alternative: There is difference in creaminess among sherbets A, B, C and D. APPENDIX E - WORKSHEETS FOR THE SENSORY EVALUATION TESTS 95 96 Table E.1. Flavor acceptance test worksheet. Type of test: Walden Sherbet A (control) Sherbet B (25% replac.) Sherbet C (50% replac.) Sherbet D (75% replac.) ”E ‘ C sssséssss Sppgpmawwu Hedonic Scale W 511 637 126 918 Mutation ABCD-511 ABDC-511 ACBD-511 ACDB-511 ADBC-511 ADCB-511 BACD-637 BADC-637 BCAD-637 BCDA-637 BDAC-637 BDCA-637 CABD-126 CADB-126 CBAD-126 CBDA-126 CDAB-126 CDBA-126 DABC-918 DACB-918 DBAC-918 DBCA-9l8 DCAB-918 DCBA-918 637 637 126 126 918 918 511 511 126 126 918 918 511 511 637 637 918 918 511 511 637 637 126 126 126 918 637 918 637 126 126 918 511 918 511 126 637 918 511 918 511 637 637 126 511 126 511 637 918 126 918 637 126 637 918 126 918 511 126 511 918 637 918 511 637 511 126 637 126 511 637 511 97 Table E.2. Texture acceptance test worksheet. Type of test: Hedonic Scale WW SW Sherbet A (control) 224 Sherbet B (25% replac.) 718 Sherbet C (50% replac.) 478 Sherbet D (75% replac.) 975 Banclistnnmher W 1,25 ABCD-224 718 478 975 2, 26 A B D C - 224 718 975 478 3,27 ACBD-224 478 718 975 4,28 ACDB-224 478 975 718 5,29 ADBC-224 975 718 478 6,30 ADCB-224 975 478 718 7,31 BACD-718 224 478 975 8,32 BADC-718 224 975 478 9,33 BCAD-718 478 224 975 10,34 BCDA-718 478 975 224 11, 35 B D A C - 718 975 224 478 12, 36 B D C A - 718 975 478 224 13, 37 C A B D - 478 224 718 975 14, 38 C A D B - 478 224 975 718 15, 39 C B A D - 478 718 224 975 16,40 CBDA-478 718 975 224 17, 41 C D A B - 478 975 224 718 18,42 CDBA-478 975 718 224 19, 43 D A B C - 975 224 718 478 20,44 DACB-975 224 478 718 21, 45 D B A C - 975 718 224 478 22, 46 D B C A - 975 718 478 224 23, 47 D C A B - 975 478 224 718 24,48 DCBA-918 126 637 224 0 Prepare questionnaires and write the panelist numbers on them. Put ques- tionnaires in right order (4 for each panelist). 0 Prepare samples. 0 Serve panelists. Make sure they taste all samples, in the right order. 0 Collect questionnaires, measure and record responses. 0 Analyze data using 2-way AN OVA. Use 'I‘ukey’s test to find which samples are significantly different. 98 Note: The steps followed for all acceptance tests were the same, only the sample code numbers were different. Table E.3.Iciness comparison test worksheet (test performed after 9 days of storage). Type of test: Schefl‘e Pair Comparison Test WW Wombat Session 1 Session 2 Session 3 Sherbet A (control) 575, 123, 377 284, 511, 212 439, 898, 779 Sherbet B (25% replac.) 392, 891, 789 686, 583, 983 167, 154, 849 Sherbet C (50% replac.) 968, 455, 254 194, 748, 834 448, 221, 633 Sherbet D (75% replac.) 289, 345, 458 414, 719, 855 435, 659, 324 Panelistnnmher W111 1 DA C-B B-D D-C A-C B-A 2 AB DB A-D C-A C-D B-C 3 B-D B-A A-C C-B A-D DC 4 C-D D—B B-C C-A D-A A-B 5 C-B A-D A—C D-C B-D AB 6 ‘ C-A B-C D-A B-A C-D D-B Emlistnnmhcr MW 1 A-C C-B B-A D-C B-D AD 2 C-A DA DE B-C C-D AB 3 DB C-D C-A C-B B-A A-D 4 B-D D-A B-C D-C A-B AC 5 A-D B-D GD GE B-A C-A 6 D-C A-B D-A B-C - D-B A-C Emelistnnmher Won!) 1 B-A C—A B-D A-D D-C C-B 2 DA B-C A-C D-B C-D A-B 3 C-A B-A C-B B-D A-D C-D 4 A-B D-C D-B B-C D-A A-C 5 DA DE DC A-C A-B B-C 6 C-A ' A-D B-D B-A C-D C-B 99 0 Prepare questionnaire, write sample codes in appropriate order and panelist number on them. 0 Put the 6 questionnaires that every panelist will take in each session in right order and staple them together. 0 Prepare samples. 0 Serve panelists. 0 Collect questionnaires, record responses. 0 Analyze data using AN OVA with and without interaction [O’Mahony (1976), Larmond (1977)]. Note 1: Three code numbers are required for each sample in each session, because all panelists taste all 6 possible pair, of sherbets, in which the same sample are presented 3 times to each panelist. So, each time the same sample is presented to a panelist, no matter if it is in the same session, or in different sessions, it should have a different code number. This way the panelist will not be biased. Note 2: All comparison tests were performed the same way, so here only the sample code numbers and the presentation order will be given for the rest of the comparison tests. 100 Table E.4. Sweetness comparison test worksheet (test performed after 10 days of storage). Type of test: Schefl'e Pair Comparison Test Smleldcnfification WW Session 1 Session 2 Session 3 Sherbet A (control) 693, 767, 688 233, 142, 857 253, 824, 128 Sherbet B (25% replac.) ’ 355, 883, 761 875, 722, 156 442, 993, 615 Sherbet C (50% replac.) 865, 549, 256 117, 733, 513 599,389,853 SherbetD (75% replac.) 542, 824, 484 937,247,197 121,866,793 Eanelistnnmber W11 1 A-B C-B B-D C-A A-D DC 2 B-C B-A D-B D-A A-C GD 3 DB D-A C-A D-C C-B B-A 4 A-C B-D D-C B—C A-D A-B 5 C-B B-D B-A A-D C-A D-C Panelistmher W 1 DA DE B-C A-C C-D A-B 2 A-C A-B D-C B-D A-D C-B 3 BA D-B D-A C-A C-D B-C 4 B-A D-A C-B D-B C-A C-D 5 AD A-B D-C B-C B-D AC 2 l’ I l 0 1 :2 I I' I . 8] 1 DC A-B B-C A-C A-D DB 2 ' D-A C-B B-A C-D B-D C-A 3 B-D A-C D-C C-B D-A A-B 4 C-D D-A B-C B-A C-A DB 5 B-A C-A C-D C-B A-D B-D 101 Table E.5. Creaminess comparison test worksheet (test performed after 11 days of storage). Type of test: Schefl'e Pair Comparison Test W Member Session 1 Session 2 Session 3 Sherbet A (control) 651, 173, 851 989, 738, 411 824, 978, 127 Sherbet B (25% replac.) 144, 377, 856 629, 365, 487 336, 164, 263 Sherbet C (50% replac.) 831, 981, 638 268, 292, 585 763, 776, 299 Sherbet D (75% replac.) 927, 838, 734 531, 972, 437 722, 924, 471 2 l' I l D l [E I l' I . ll 1 BC C-D D-A A-C B-D B-A 2 GB DE DC C-A A-B A-D 3 DC C-A A-B C-B D-B A-D 4 A-C B-A D-A B-D B-C C-D 5 C-B C-A D-C D-B D-A B-A Panelistnnmher W21 1 B-D C-D A-D A-C B-C AB 2 C-A D—C B—D D-A B-C B-A 3 A-B D—B A-D A-C C-B C-D 4 DC B-D B-C B-A D-A C-A 5 . D-B A-C A-D C-D C-B A-B Baneliatnnmher W13) 1 B-A A-C D-B B-C D-A DC 2 GA C-D A-D A—B C-B B-D 3 C-D C-A D-B B-C D-A AB 4 A-D B-D C-B A-C B-A DC 5 DC A-B C-B D-A B-D C-A 102 Table 13.6. Iciness comparison test worksheet (test performed after a heat shock treatment). Type of test: Schefl'e Pair Comparison Test Widen WW Session 1 ‘ Session 2 Session 3 SherbetA(control) 313, 572, 890 141, 227, 695 885, 714, 519 SherbetB (25% replac.) 691, 418, 225 775,137,632 869, 323, 448 Sherbet C (50% replac.) 979, 622, 544 719,148,242 912, 827, 213 SherbetD (75% replac.) 375, 785, 449 981,263,699 240, 834,578 Eanelistnnmbsr W11.) 1 DA B-C A-C DC DE A-B 2 A-D C-D B-D B-A C-A C-B 3 B-C B-A D-B D-A A-C C-D 4 A-B B-D A—D D-C C-A C-B 5 A-D C-A C-B C-D A-B B-D 6 DA B-A D-C A-C B-C D-B Panelistnnmher W21 1 C-A B-C D-A B-D D-C B-A~ 2 C-B A-C A-D C-D A-B DB 3 DB A-C B-A D-A C-B DC 4 C-A B-C B—D C-D A-B A-D 5 C-B A-D B-A DE DC A-C 6 D-A B-C C-A C-D A-B B-D Emelistnnmhsr W13.) 1 A-D A-B D-B C-D A-C C-B 2 B-C B-D D-C C-A B-A DA 3 B-C C-A D-C A-B B-D DA 4 A—C B-A C-B C-D A-D DB 5 D-C D-B A-C A-D A-B B-C 6 DA C-D C-B C-A B-D B-A 103 Table E.7. Iciness comparison test worksheet (test performed after 32 days of storage). Type of test: Schefl'e Pair Comparison Test Samaleldcntifismion Samlafledennmher Session 1 Session 2 Session 3 Sherbet A (control) 392, 968, 289 214, 884, 705 696, 811, 115 Sherbet B (25% replac.) Sherbet C (50% replac.) Sherbet D (75% replac.) 575, 121, 470 129, 258, 607 180, 316, 926 837, 716, 931 356, 101, 448 621, 504, 730 309, 662, 583 167, 127, 566 113, 239, 521 01:2 ||°(°ll * The sixth panelist did not have the test, so the last row in each session was not used. B-C D-C A-C B-D C-B C-D A-D A-B D-B C-A D-C A-D C-D D-A C-B B-C A—B D-B A-C C-A D-A C-D B-D B-A D-B . B-D D-C A-D C-A A-C B-A C-B B-A A-B D-A B-C 01:2 ll'('2] A-B D-C B-A A-B C-D D-C C-B B—A D-A C-B D-B B-D A-C D-B B-C A-D B-C A—D C-D B-C A-C D-C B-A A-B D-A C-A D-B C-A C-A A-C B-D A-D’ C-D B-D D-A C-B 01:2 ll'l'fll B-D B-C A—D B-D C-D D-A C-B D-C C-B D-A B-C C-A A-B B-A C-A B-A DE DC C-D D-B D-B A-C A-C B-D A-D D-A A-B B-C A-B C-B C-A A-C C-D D-C A-D B-A APPENDIX F VERBATIM FROM THE SENSORY EVALUATION TESTS 'Verbatim for the flavor of sherbet A (control) 0 The flavor of the sample was not overpowering. 0 Very good. Tastes great. 0 Tastes fruity, but not very “orangy”. 0 It does not seem to have the strong orange sherbet flavor. ' Seems creamier. ' Not very tangy. ° This is much too sweet. Not enough orange flavor. 0 More milky/chalky quality. 0 Mouthfeel is good, but it tastes too sweet. 0 A strong orange flavor and a good afiertaste. I place it at the top of the list. 0 It needs a little more sweetness. 0 Flour like sensation in mouth. Little flavor. ° Creamier than a sherbet should be. 0 Too sweet, nice odor. 0 Tastes rather bland. I might like a stronger flavor. 0 It was not as tangy as most orange sherbets I have tasted. Tasted almost like ice cream. ' I liked it, because it had no strong aftertaste. ‘ I normally don’t like orange sherbet, but this one does not have too strong of an orange taste, it does not taste like baby aspirin. 0 Tasted like a candy. O Fruity, but seems to have a slight aftertaste. 0 Did not taste very orangy. It tasted more than eating cream than orange. 0 It does not taste like orange sherbet to me. 0 A little dull, but tasted very good. 0 Not real orange taste. 0 Could be more fruity. 104 105 0 Had good taste. Noticed a slight sweet aftertaste, like that from a sugar sub- stitute. O Did not taste like orange, tasted like artificial flavor. ° Not as fresh tasting (comment after storage for 122 days). ' Slightly heavy taste. ' Very tasty. Like a creamsicle. 0 Excellent flavor. 0 Sweet with a lot of orange taste. 0 Does not taste like sherbet, too smooth. ' Just the right creaminess, but needs a touch more orange. . 0 Taste was average, but it left a coated type feeling in my mouth. 0 Most natural, good lingering fruit flavor. 0 Great taste, somewhat chewable. 106 0 Verbatim for the texture of sherbet A (control) 0 The sample was quite smooth. 0 Very creamy and smooth. 0 Flat texture. ‘ Creamy, pleasant texture. 0 Not as hard as sherbet C. ' Good texture. 0 Smoother than usual sherbets. ' The texture was excellent. There was no ice in it. It was very creamy. ' Creamy, almost chewy. 0 Too creamy for a sherbet. ° Spoons well. ' Soft texture. ' There was hardly any texture, it melted too quickly in my mouth. A cone would be gone too fast! 0 Felt natural. 0 Too soft for sherbet. 0 It is just right. It does not melt too easily. 0 It is a bit dry tasting, almost like it was in the freezer for a while. ' Too smooth. I like it with an icier or less smooth texture. Comments after the heat shock treatment 0 It was totally creamy and smooth. 0 Texture very smooth. Not much effort for dissolving or swallowing. 0 Smooth, but could be improved. 0 More watery, icy. ' More like ice cream than sherbet. 0 Nice mouthfeel, did not melt too quickly. 0 Many ice crystals. I would prefer it a bit more harder. 107 ' Verbatim for the flavor of sherbet B s This sample had a very light orange flavor, almost creamy flavor, it was very good. 0 It had a very sweet taste, while melting in my mouth. ' Some aftertaste, but not very strong. ' Good flavor, not too acidic. 0 It has the right amount of sherbet flavor, yet not too strong. 0 Not very tangy. 0 Nice orange flavor. Slightly too sweet, but good. ' Strange flavor. 0 Has more milky/chalky flavor. The orange flavor seems enhanced. 0 Strong milky taste. 0 Tastes good, but it almost has too much creamy taste. It tastes too watery. ' Good mouthfeel. Great tangy creamy taste! 9 It tastes like real oranges. 0 I would prefer it with more orange flavor. 0 It would have been extremely good if the taste of orange was more dominant. ' Chalky aftertaste. ' Gummy flavor. 0 Too sweet. 0 A little creamy aftertaste. 0 More orange flavor than sherbet A. 0 Tangier than sherbet A. ° Tastes like sherbet, not ice cream. Creamy. 0 Not enough orange flavor or sugar. It also left an unpleasant aftertaste. ' It has a sort ofa bland taste. 0 Very creamy in mouth, tastes good. 0 It is not too creamy. It has the right amount of sweetness. 0 Left an aftertaste in mouth. 0 It was rather watery. 0 Very creamy, aftertaste, does not taste as naturally flavored. ' Fruity and creamy. 0 Mild flavor. 0 It was creamy and tasted like a creamsicle. 0 I liked it, however an orange sherbet should have stronger flavor. O The initial taste was great, but it seemed after it dissolved that there was an aftertaste like medicine. It took away the initial pleasure. ‘ Very nice and light flavor. ' Not a bad aftertaste and tasted very well. 0 Sort of dull. 0 It tastes more sherbet, did not leave any aftertaste, but contained a lot of sugar. 108 0 Basically tasted like regular, but good orange sherbet. 0 Good strong flavor, no aftertaste. 0 Flavor did not have a long lasting power. ' Tasted a little like cream cheese. ' Very creamy. Much too sweet. More sugary than orangey. 0 Good, a lot of orange taste. 0 Pleasant aftertaste. ' Flavor seemed to last. ' Creamy, good mouthfeel, good flavor. ' It tastes as if cheese was added to it. 0 I don’t know why I like it so much, but it just tastes like the sherbet you get at the restaurant. 0 Slight ofl‘ flavor. Very bland in orange flavor. 0 It tastes little funny. 0 I liked the very faint, but noticeable taste. 0 Hardly any taste at all. 0 It was not too sweet, did not leave any bad aftertaste. Made me want more. 109 ' Verbatim for the texture of sherbet B 0 Smooth. Almost silky. ' I like the creamy texture, as opposed to a more icy texture that sherbet usu- ally has. 0 Ice crystals present. 0 Good texture. 0 It is soft and melts in the mouth. I enjoyed it a lot. ’ Gummy texture. O I like the consistency of this sample. 0 More like ice cream texture. 0 Very similar to sherbet A. 0 Not too rough, not too smooth, just perfect. 0 It is too creamy and rich for sherbet. 0 Spoons well. ' Soft and smooth texture. 0 Seemed to be smooth, but broke down in small chunks. 0 Firm texture. 0 Seems a little gluey. 0 Seems too soft for sherbet. 0 It is too smooth, I like it with a harsher texture. Comments after the heat shock treatment 0 Texture is somewhat rough. 0 Smooth, but a little hard. 0 Too crunchy. 0 A slight resistance to melt. Firm texture on tongue. 0 This sample was excellent. Not too 8011: not too hard. 0 Almost too soft. 0 Needs to be a little stiffer. 0 Not enough body. 0 It seems smoother, less crystalline than usual sherbets. ' Somewhat gummy. 0 Iciness and hardness O.K. 110 ' Verbatim for the flavor of sherbet C 0 Although I liked it, the flavor was quite strong. 0 Didn’t taste very orangy. Tasted like cream cheese. ' Buttery taste. Left an aftertaste. 0 Flavor is appropriately strong. 0 Flavor is off. 0 Good, but not so much orange flavor. 0 Very good flavor. 0 Not very tangy. 0 This has a slightly sweet aftertaste. Good flavor. 0 Too sweet for me. Good strong orange flavor, though. 0 There is a strange feeling after I ate the sherbet. ' Weird aftertaste. Not much orange flavor. 0 It is not sweet enough, it has too much cream or milk, it tastes filmy. ' 0 Good mouthfeel, less creamy, but very good tangy taste. Does not taste too sweet, which is the way I like sherbet. This one was my favorite. 0 Too creamy. 0 Chalky aftertaste and too strong of a flavor and texture. 0 Creamy aftertaste, strong. 0 Tastes slightly sour. Sweetness OK. 0 Has profound creamy flavor, outpowers the orange flavor. 0 Needs stronger flavor. 0 Tangier than sherbet A. 0 It has a pretty good flavor, it is not bitter, but it has too strong of an orange flavor. 0 Tasted terrible! 0 Not very flavorful. ' There is no distinct flavor. 0 Flavor was a little ofl‘. 0 Fruity and sweet. 0 Weird aftertaste. 0 I liked it, but it was sweet. 0 .Very good flavor. I would eat this anytime it was offered. 0 Vanilla taste to it, did not enjoy the aftertaste. 0 Too creamyandhasadryaftertaste. 0 Orange flavoring not right. 0 Had a nice creamy orange flavor to it. ' This has a very strong orange flavor. It is not very sweet and this is good, because there is more flavor and less sweetness. ' Very creamy for sherbet, which I liked. 0 Nice, rich flavor. 0 Very tangy flavor, it has a good mouthfeel. 111 0 Tasted a lot like orange flavored cream cheese. 0 Extremely good. 0 Has cream cheese flavor. Lacks orange. 0 Not too sweet and not too creamy. ' Seems watery and light. Like popsicle. 0 More refreshing tangy taste. 0 Very odd of flavor, when first put in month. Also, lacks orange flavor. 0 Strong taste at first, but then no taste. 0 Stronger taste of something other than orange. ' Good orange flavor. 0 Too sweet. Left a weird taste in my mouth. 0 Less fruity than D. 0 Almost a bitter aftertaste. 112 ' Verbatim for the texture of sherbet C 0 Good texture. 0 Much more icy, therefore crunchy. ' Not as smooth as it should be. 0 Although it was a little too soft, it did not melt in the mouth. Liked it, but not very much. 0 Hard. 0 Texture not as nice as for sherbet A. 0 Tasted like fi-ozen orange yogurt. 0 It holds together in the mouth, which gives you a rich and smooth feeling. 0 Harder than sherbet B. ' A little bit grainy. 0 Icy consistency. . 0 It was not soft enough, it was too hard. * 0 Slightly icy. ' Icy. grainy- . A little harder than sherbet A. ' Texture good for sherbet. 0 A little bit coarse texture. 0 Too hard, not smooth. ' Crystals were a bit too large. ' Very creamy (more so than most sherbets). 0 Spoon went in hard. 0 Does not melt in mouth like sherbet does. 0 I would have liked a slightly firmer texture. 0 Harder than sherbets A and B. ' A little too crystal like or icy. 0 Very clumpy. 0 Too thick. 0 Kind of chalky. 0 Crispy. 0 Chunky. Did not melt smoothly. 0 Leaves a sort of dry feeling. Comments after the heat shock treatment ' Smooth texture. Not icy. 0 The texture was just right. Easy to swallow. 0 Nice and smooth. ' Smooth texture. 0 A little bit rough and harder than sherbet B. 113 0 This is the best texture. It is icy, yet creamy enough like normal sherbet. ' Semi grainy. ° Grainy aftertaste. 0 Very smooth. 0 Clean icy melting in mouth. 0 Crumbly texture. Not a smooth meltdown. 0 Too watery or slippery. 0 Could melt in my mouth a little more. ' Slightly gritty. 0 A little clumpy. ' Creamy, but not fatty texture. 0 Very hard, crumbling. 114 O Verbatim for the flavor of sherbet D O I think the sample’s flavor is lacking something. O Tasted like cream cheese, instead of orange sherbet. O Strong buttery taste. Does not taste like sherbet. O Sample had an unusual aftertaste. O First taste (initial impression) was highly favorable. But an aftertaste lingers which is not good. The smell of the sample was inferior. O Not enough orange flavor comes through. O It has more of a sour cream undertone. O It seemed to have a funny taste to it. Not bad, but funny. Less sweet. O Good orange flavor. Could be sweeter. O The first taste is bitter. then you taste nothing. O I am not sure why I dislike it. Strange flavor. O Tasted good and not as tart as some I’ve had before. O It has an aftertaste like cream cheese. O Mouthfeel is good. Taste is tangy (good taste!). O Taste is very good, strong orange flavor. Aftertaste is milky, chalky. It is not too sweet, and this is good. O Tastes too sour. O Too creamy. O Not too sweet. It has a kind of cream cheese taste. O The worst of all. O More vanilla taste. O It had a lemon like flavor, that I did not enjoy. O Good flavor. O Tastes a little too creamy. O Left a chalky aftertaste, and after melting it seemed like a residual was left behind. O Orange flavor not very intense. O The sweetness is just great. Tastes slightly sour (greatll). Odor is just enough. O It really almost tastes like cheese cake. It tastes rich and creamy. O Rich flavor. O Tasted good, sweet, tangy mouthfeel. O A little too bland. It did not taste much like orange. - It has a bitter taste to it, it tastes like baby aspirin. O Terrible. O Has a slight cream cheese taste to it. O Tastes like sour milk. O Has a distinct orange flavor, but not strong enough. O Nice feel on mouth. O Strange flavor. O Of all samples, I disliked this one the most. Although it was fruity as sherbet 115 should be, it had a different flavor to it, that I disliked. O Bad aftertaste, funny flavor. O It tasted orange, but is probably not something I would eat regularly. There was not much lingering flavor after it was gone. O I liked the product a lot, but the aftertaste was different than the product. O I enjoyed the sweetness of the sample along with the orange flavor. O Good stuff. O Too heavy. The consistency is like fi‘ozen yogurt. O Not as creamy. O Did not taste like orange. Had a had first impression. O Taste is good, except a little bit sour. O I really liked the flavor, it tasted like oranges, but it was not too potent. O the orange flavor did not taste right, but I can’t pick out what it was. ' Chalky taste to it. O It left a distinct aftertaste in my mouth. O Had a sort of creamy flavor to it, not really distinct. O Tasted sour, a lot like cream cheese. O The first taste was a slightly cream cheese flavor. This does not have much of a lingering aftertaste (which is good). O At first it does not quite taste completely like orange, but then it does. O Tasted like custard, not orange sherbet. O Too milky. O Tastes more like ice cream. O A little bit tasteless. I would like more flavor. O More refi'eshing, tangy, sherbet taste. O It tasted the best of all. O Very strong off flavor. Tastes like orangey cottage cheese. O It has an off tangy flavor. O Pretty much perfect. O Has a harsher flavor. Seems more acidic. O Needs more orange flavor. O It did not really taste the way sherbets usually taste. O Too sweet. As soon as I put it in my mouth I did not like the taste. It was unpleasant. O Very strong, almost spoiled aftertaste. O Very sweet, cool and fruity. 116 ' Verbatim for the texture of sherbet D O Some ice crystals. O Texture is too smooth, nothing to chew. O A little less icy would be better. O Not as smooth. O Not as soft as the sherbet A, but still good. O Very smooth. O Not nice texture. O Tastes buttery. O It tears apart in dish. O Good texture. O A little gritty. O King of gummy. O Very gritty. O Funny, awkward texture. O A little hard. O More body than most sherbets. O Odd texture. O Seemed to feel rough to the tongue and roof of mouth. O It would be especially good for use on a cone. It is firm enough. O More crystally texture. Very similar to sherbet C. O Chalky feel, coats mouth with chalk. O Sort of a gritty or icy tasting at first, but then after it has been in mouth for a few seconds, it has a nice texture. O Too many ice crystals, not smooth enough. O Melts away too fast, does not last in your mouth. O Dissolve way too fast. Clump. O Harder than the others. O I prefer the texture of this sample. It was more icy. O Harsh, more frozen. O It broke apart and was very icy and rough to my tongue. O Nice and creamy, but it leaves a sticky, gross aftertaste. O, Very nice texture, although a little too slippery. O Crunchy. O Was not bad, but it had lumps. It didn’t seem to melt in your mouth as easy as the others. O It has got a nice icy texture to it. Because it is not extremely smooth. O After I swallowed it, it left a sort of weird dry sensation in my mouth. 117 Comments after the heat shock treatment O It seems to be dry and crumbly. Slightly grainy. O Seemed icy, not very smooth. O This sample was much too rough. O Very gritty and hard. O Did not hold together. O Clean, not grainy, more resistant to melting or firmer consistency than sher- bet C. O Texture very good at first, but seemed floury afterwards. O Starts very icy, but then melts smoothly in your mouth. O Slight fine light texture. O Did not melt quite as easily as I expect sherbet to melt. O Too hard on top, watery, coats tongue. O I would prefer it less icy. APPENDIX G RHEOLOGICAL RAW DATA (SHEAR STRESS / SHEAR RATE) FOR THE SHERBET MIXES 118 119 Table G.1. Shear rate and shear stress data for sherbet mixes at 40°F (4.44°C). ' number Replication Sherbet A 7.694 13.88 25.50 40.18 58.03 81.08 116.9 165.2 211.9 7.859 14.06 23.26 47.10 70.60 117.4 164.8 235.2 8.202 - 14.11 24.35 36.63 58.23 81.22 116.6 164.4 210.7 0' 0.4711 0.6779 1.174 0.9883 1.571 2.517 3.716 5.021 6.087 0.4711 0.6002 0.9846 1.755 2.431 3.906 5.224 7.205 0.4711 0.5674 0.9573 0.938 1.579 2.279 3.622 4.797 6.181 9.612 14.36 23.36 35.18 58.22 81.52 117.0 163.7 209.9 7.829 14.17 23.26 35.49 58.88 81.53 117.1 164.6 235.7 8.051 23.99 35.60 58.61 81.60 116.6 163.8 234.9 0’ 0.4711 0.5144 0.8187 1.129 1.782 2.699 3.822 5.393 6.966 0.4711 0.5731 0.9642 1.409 1.871 2.718 3.986 5.365 7.550 0.5521 1.026 1.304 1.976 2.792 4.178 5.898 8.511 13.33 16.73 23.04 35.07 58.90 81.52 116.7 164.2 211.7 7.874 19.00 35.11 58.91 82.15 117.1 164.0 - 210.4 16.24 35.67 58.45 81.09 117.0 165.3 211.1 0' 0.4711 0.4957 0.7302 1.198 1.619 2.346 3.479 ' 4.904 6.171 0.4711 0.6332 1.192 1.802 2.402 3.472 4.748 6.161 0.4711 0.962 1.215 2.020 2.932 4.073 4.887 10.12 19.50 42.86 58.36 91.50 138.5 185.7 233.4 14.67 23.79 35.21 58.23 80.80 115.8 162.9 211.9 12.42 23.51 35.22 58.46 81.04 116.3 163.9 211.2 0.4711 0.5191 0.6317 0.6655 1.644 3.047 4.434 5.718 0.4711 0.5224 0.8243 1.090 1.857 2.841 4.503 5.706 0.4711 0.6077 0.9803 1.274 2.044 3.089 4.527 5.670 APPENDIX H AN OVA TABLES FOR OBJECTIVE AND SENSORY TESTS Table H.1. AN OVA table for the viscosities of sherbet mixes. satiation freedom mates Between 3 84.76 28.25 665* Within 8 33.97 4.25 Total 11 118.73 * Significant difi‘erence at p<0.05. Table 11.2. AN OVA table for the melting resistance of sherbets at 38°C (100.4°C). WWW Watchman Wheaten m Sherbet, A 3 11142.906 3714.302 97.4129* Error 4 152.518 38.129 Time, B 27 354980.121 13147.412 1788.3532 AB 81 4786.219 59.089 8.0375 Error 108 793.982 7.352 Total 223 371855.746 * Significant difference at p<0.001. 120 121 Table H.3. AN OVA table for the flavor acceptance of sherbets after 8 days of storage. Sources! W Sunni. Mommas: Elaine inflation harden mam Judge 47* 407.92 8.68 4.35 Sample 3 46.83 15.61 7.83" Error 141 281.17 1.99 Total 191 735.92 "' 48 judges tested, all responded. “ Significant difi'erence at p<0.001. Table HA. AN OVA table for the texture acceptance of sherbets after 8 days of storage. Scum W Sunni. Meansmnre E-xalne satiation freedom mam judge 47* 211.33 4.496 2.12 sherbet 3 65.18 21.727 10.26” Error 141 298.57 2.118 Total 191 575.08 "‘ 48 judges tested, all responded. *" Significant difference at p<0.001. Table H.5. AN OVA with interaction table for the iciness test after 9 days of storage. Source new Sunni Monuments 121181119 2111811811211 harden mam Sherbet (a) 5 68.00 13.60 16.88 Judge (A) 5 16.78 3.36 4.17 axA 25 24.22 0.97 120* Error 72 58.00 0.81 Total 107 167 * No significant sherbet-judge interaction. 122 Table H.6. AN OVA table for the Scheffe paired comparison test for the iciness of sherbets after 9 days of storage. Sourcesli nemssni Sunni Mommas Emilie satiation Medan mates Main efl‘ect (sherbet) 3 38.84 12.95 6527* Order efl‘ect 1 1.836 1.836 0.925“ Error 103 204.324 1.984 Total 107 245 " Significant difference among sherbets at p<0.001. “ No significant order effect. Table H.7. AN OVA table for the Schefie paired comparison test for the sweetness of sherbets. Snurceni Dammni Sunni. Memento E-nlue sedation burden mm Main effect (sherbet) 3 67.416 22.472 15.236* Order effect 1 0.216 0.216 0.147** Error 85 125.368 1.475 Total 89 193 * Significant difference among sherbets at p<0.001. ** No significant order effect. 123 Table H.8. AN OVA table for the Scheffe paired comparison test for the creaminess of sherbets. SnumeniDnmniSunniMnnusummalue Malina tendon mam Main efl‘ect (sherbet) 3 113.47 Order efl'ect 1 1.475 Error 85 98.06 Total 89 213 * Significant difference among sherbets at p<0.001. ** No significant order effect. 37.82 32.77“ 1.475 1.274“ 1.154 Table 11.9. AN OVA table for the flavor acceptance of sherbets after a heat shock treatment. Snumninemni undefinufizendnn Judge 47* Sherbet 3 Error 141 Total 191 "' 48 judges tested, all responded. ** Significant difi‘erence at p<0.01. Sunni. scum 215.83 - 36.93 426.32 679.08 Wanna 4.592 1.52 12.311 4.07.“ 3.024 Table 1110. AN OVA table for the texture acceptance of sherbets after a heat shock treatment. Snumninemm undefinuimednn Judge 47"I Sherbet 3 Error . 141 Total ‘ 191 "' 48 judges tested, all responded. “ Significant difference at p<0.001. Sunni. m 207.74 84.81 310.44 602.99 Wit-nine 4.420 2.01 28.269 12.84“ 2.202 124 Table H.1 1. AN OVA table for the Schefle paired comparison test for the iciness of sherbets after a heat shock treatment. Snumni Remnant SunniMeausuuaranzalue Malina freedom sum Main efl‘ect (sherbet) 3 66.22 22.07 18.39* Order efl‘ect 1 0.15 0.15 0.124" Error 103 123.63 1.20 Total 107 190 * Significant difference among sherbets at p<0.001. ** No significant order effect. Table H.12. AN OVA table for the flavor acceptance of sherbets after 31 days of storage. SnumninemmniSunnLMeaunnuamEnnlue uriafinu inaction mum Judge 47"l 331.00 7.043 2.42 Sherbet 3 25.08 8.361 2.87“ Error 141 410.92 2.914 Total 191 767.00 * 48 judges tested, all responded. ** Significant difference at p<0.05. Table H.13. AN OVA table for the flavor acceptance of sherbets after 122 days of storage. satiatiniirendnn squares Judge 47* 219.74 4.675 1.52 Sherbet 3 59.35 19.783 6.42“ Error 141 434.40 3.081 Total 191 713.49 * 48 judges tested, all responded. ** Significant difference at p<0.001. 125 Table H.14. AN OVA table for the texture acceptance of sherbets after 31 days of storage. Snurnnninemnani Sunni. Meannnuarcr‘nnlue anxieties finndnn scum Judge 47* 238.17 5.067 1.94 Sherbet 3 66.42 22.139 8.46** Error 141 369.08 2.618 Total 191 673.67 "' 48 judges tested, all responded. ** Significant difference at p<0.001 Table H.15. AN OVA table for the Schefl'e paired comparison test for the iciness of sherbets after 32 days of storage. SnutcaniWSunniMmssuanmue Malina irnndnn scum Main effect (sherbet) 3 94.51 31.50 59.22“ Order effect 1 2.28 2.28 . 4.29“ Error 85 45.21 0.53 Total 89*** 142 * Significant difference among sherbets at p<0.001. ** Significant order effect at p<0.05. *** 5 trained panelists participated in this test. BIBLIOGRAPHY Chapter 6 BIBLIOGRAPHY AOAC. Assosiation of Official Analytical Chemists. 1990. Oficial Methods of Analysis. Vol. 2. Helrich, K. (editor). 15th ed. Arlington, VI. pp. 805, 808, 811, 812, 831, 851. ADPI. American Daily Products Institute. 1991. Information Pack. Anon. 1976. Quality frozen desserts... and a savings of 12c per gallon. Food Proc. 7:49. Anon. 1979. Using modified whey in dairy products. Dairy & Ice Cream Field. 162 (4):82I. Anon. 1981. Membrane filtration gets a second look. Food Engin. 53(7):60. Arbuckle, W. S. 1979. Whey solids in frozen dessert formulations. American Dairy Review 41(12):50D. Arnold, R. G., Evans, T. A., Kreshel, C. L. 1976. Dairy & Ice Cream Field. 159(11):55. Babella, G. 1984. The development and utilization of milk and whey-protein concentrates in Hungary. Developments in Food Sci. 9:241. Bodyfelt, F. W., Matthews, M. V., Morgan, M. E. 1979. Flavors associated with the use of Cheddar cheese whey powder in ice cream mix. J. Dairy Sci. 62 (Suppl. 1):51. Carter, C. A., Ashton, G. 0., Pearson, A. M. 1982. The influence of ingredients on the firmness of ice cream. Modern Dairy. 61(5):19. 126 127 CPR Code of Federal Regulations. 1990a. Title 21 - Food and Drugs. Part 184- Direct Food Substances Afirmed as Generally Mgnized as Safe. Subpart B - Listing of specific substances ailirmed as GRAS. Sect. 184.1979 - Whey. Fed. Reg. Washington DC. CFR. Code of Federal Regulations. 1990b. Title 21 - Food and Drugs. Part 135- Frozen Desserts. Subpart B - Requirements for specific standardized frozen desserts. Sect. 135.110 - Ice cream and frozen custard. Part 3b - Optional ingredients. Fed. Reg. Washington DC. CFR. Code of Federal Regulations. 1990c. Title 21 - Food and Drugs. Part 135- Frozen Desserts. Subpart B - Requirements for specific standardized frozen desserts. Sect. 135.140 - Sherbet. Fed. Reg. Washington DC. Christensen, V. W. 197 6. Whey utilization in the United States. Dairy Ind. Intern. 41(3):84. Coder, D., Parsons, J. G. 197 9. The efl‘ects of processed wheys and caseinate on composition and consumer acceptance of ice cream. J. Dairy Sci. 62(Suppl. 1):35. Cooper, T. G. 1977. The tools of biochemistry. A Wiley-Interscience publication. J. Wiley & Sons, Inc. New York. Crocco, S. C. 1975. Ultrafiltration excels in whey protein recovery. Food Engin. 47(11):59. Crowe, L. K. 1960. Results obtained with a panel preference evaluation of ice cream. Ice Cream Field. 75:19. Demott, B. J ., Sanders, 0. G. 1980. Use of direct-acid-set Cottage cheese whey to manufacture sherbet. J. Food Protec. 43(6):433. Farrell, H. M. Jr, Douglas, F. W. Jr. 1983. Effect of the UHT pasteurization on the fimctional and nutritional properties of the milk proteins. Sympo- sium on role of milk proteins in human nutrition. Heler Milch. Forsch. 35:239. Frazeur, D. R. 1959. Factors affecting the churning of ice cream. Part III: Mis- callaneous factors. Ice Cream Field. 73(7):48. 128 Frazeur, D. R., Harrington, R. B. 1967. Consumer preference for frozen des- serts containing wheys. Ice Cream Field & Trade J. 149(6):40. Frazeur, D. R. 1967. The use of wheys in frozen desserts. Ice Cream Field & Trade J. 149(5):22. Frazeur, D. R. 1977. Ice cream: keeping up with changes. Dairy & Ice Cream Field. 160(5):62. Glass, L., Hedrick, T. I. 1977a. Nutritional composition of sweet- and acid- type dry wheys. I. Mnjor factors including amino acids. J. Dairy Sci. 60(2):185. Glass, L., Hedrick, T. I. 1977b. Nutritional composition of sweet-and acid- type dry wheys. 11. Vitamin, mineral and calorie contents. J. Dairy Sci. 60(2):190. Goldsmith, R. L. 1981. Ultrafiltration production of whey protein concen- trates. Daily field. 164(8): 88. Guy, E. J. 1978. Partial replacement of non-fat milk solids and cane sugar in ice cream with hydrolyzed sweet whey solids. [Abstract]. J. Dairy Sci. 61(Suppl.l):106. ‘ Hansen, P. S., Jensen, G. K. 1977. Investigations concerning variations in the composition of whey. 224. Pages 1-77 inAnnual Report of the Statens Forsogsmeieri Hillerod. Denmark. Hansen, A. P. 1979. Using whey in ice cream. Daily & Ice Cream Field. 162(7):53. Hekmati, M., Bradley, R. L., JR. 1979. Savings with acid whey in sherbet manufacture. Dairy & Ice cream Field. 162 (2):66. Holmes, D. G. 1979. Whey products. New Zealand J. Dairy Sci. & Technol. 14(2):208. Horton, B. C., Goldsmith, R. L., 2811 R. R. 1972. Membrane processing of cheese whey reaches commercial scale. Food Technol. 26:30. Huse, P. A., Towler, 0., Harper, W. J. 1984. Substitution of non-fat milk solids in ice cream with whey protein concentrate and hydrolyzed lactose. New Zealand J. Dairy Sci. & Technol. 19(3):255. 129 IAICM. 1965. Production index of ice cream and related products, 1964. Spec. Bull 110. Int. Assoc. Ice Cream Mfrs., Washington, DC. IAICM. 1984. The latest Scoop. Int. Assoc. Ice Cream Mfrs., Washington, DC. Igoe, R. S., Watrous, G. H., Keeney, P. G., MacNeil, J. H. 1973. Utilization of Cottage cheese whey in ice cream. Dairy Ice Cream Field. 156(5):61. J elen, P. 1979. Industrial whey processing technology: an overview. J. Agric. & Food Chem. 27(4):658. J elen, P. 1983. Reprocessing of whey and other dairy wastes for use as food ingredients. Food Technol. 37(2):81. Kock Membrane Systems Inc. Operating Manual for the $4 Spiralwound Ul- trafiltration pilot plant. Wilmington. Massachussets. Larmond, E. 1987. Laboratory methods for sensory evaluation of foods. Agri- culture Canada. Public. 1637/E. Leighton, A. J uanary 1944. Use of whey solids in ice cream and sherbets. Ice Cream Rev. 27(6):18. Loewenstein, M. 1975. Using milk solids non-fat in ice cream. Dairy & Ice Cream Field. 158(6):42. Marshall, K. R. 1982. Industrial isolation of milk proteins: whey proteins. Pag- es 339— 373 in Developments in Dairy Chemistry. Vol 1: Proteins. Fox, P. F. (editor). Applied Science Publishers, N.York. . Martinez, S. B., Speckman, R. A. 1988. B-galactosidase treatment of frozen dairy product mixes containing whey. J. Dairy Sci. 71(4):893. Mathur, B. N., Shahani, K. M. 1979. Use of total whey constituents for human food. J. Dairy Sci. 62(1):99. Matthews, M. E., Doughty, R. K., Short, J. L. 1978. Pretreatment of acid casein wheys to improve processing rates in ultrafiltration. New Zealand J. Dairy Sci. & Technol. 13(4):216. McGugan, W. A., Larmond, E., Emmons, D. B. 1979. Some observations on the flavor of acid whey. Canadian Ins. Food Sci. & Technol. J. 12(1):32. 130 Meilgaard, M., Ceville, G. V., Car, B. T. 1987a. Sensory evaluation techniques. Vol 1. CRC Press, Inc. Boca Raton. Meilgaard, M., Ceville, G. V., Car, B. T. 1987b. Sensory evaluation techniques. V012. CRC Press, Inc. Boca Raton. Morr, C. V., Swenson, P. E., Richter, R. L. 1973. Functional characteristics of whey protein concentrates. J. Food Sci. 38:324. Morr, C. V. 1982. Functional properties of milk proteins. Pages 384-393 in Developments in Dairy Chemistry. Vol 1: Proteins. Fox, P. F. (editor). Applied Science Publishers, N.York. Muller, L. L. 1976. Whey utilization in Australia. Austr. J. Dairy Technol. 31(3):92. Newlander, J. A., Atherton, H. V. 1964. The chemistry and testing of dairy products. 3rd edition. Olsen publishing Co. Milwaukee, WI. Olson, NP. 1979. Using whey - some recent techniques. Dairy & Ice Cream Field. 162(7):55D. O’Mahony, M. 1986. Sensory evaluation of food: Statistical methods and proce- dures. Marcel Dekker., Inc. N. York & Basel. Ott, D. B. 1990. Sensory assessment of foods. Lecture and laboratory outlines for the course. HNF 310. FSHN Department. MSU. Ott, D. B. 1992. Personal Communication. Parsons, J. G., Dybing, S. T., Coder, D. S., Spurgeon, K. R., Seas, S. W. 1985. Acceptability of ice cream made with processed wheys and sodium caseinate. J. Dairy Sci. 68(11):2880. Partridge, J. A. 1983. The effects of lactic acid culture seeding of raw milk in the yield and quality of Cheddar cheese. PhD dissertation. FSHN Dept. MSU. Patel, A. R., Harper, W. J. 1977. Utilization of acid whey in ice cream. Dairy & Ice Cream Field. 160(6):7OA. Petersen, R. G. 1985. Design and analysis of experiments. Marcel Dekker, Inc. N. York. 131 Potter, F. E., Williams, D. H. 1949. Use of whey in sherbet. Ice Cream Rev. 32:44. Reid, w. H. E., Shaffer, L. o. 1947. The use of dehydrated whey solids in the manufacture of difl'erent flavored ice creams. Pages 6-13 of Vol. 2 in Proc. 43rd Ann. Conv. Int. Assn. of Ice Cream Mfrs. Richter, R. L. 1983. Exploring ultrafiltration frontiers.Dairy Field. 166(4):62. Rosenberger, W. S., N ielsen,V. H. 1955. Spray dried whey powder in ice cream mixes. Am. Milk Rev. 17(6):50. Roualeyn, I., Fenton M., Charles G. H., JR. 1971. Use of ultrafiltration/re- verse osmosis for the concentration and fractionation of whey. J. Food Sci. 36:14. Ryder, D. N. 1980. Economic considerations of whey processing. J. of the Soci- ety of Dairy Technol. 33(2):73. Saal, H. 197 6. Kraft seeks improved uses for whey and finds new applications. American Dairy Review. 38(9): 14. Sherman, P. 1988. The sensory-rheological interface. Pages 417-431 in Food Structure: its creation and evaluation. Blanshard, J. M.V., Mitchell, J. R. (editors). Butterworths, Boston. Sienkiewicz, T., Riedel, C. 1990. Whey and whey utilization. 2nd ed. Verlag Th. Mann, Gelsenkirchen-Buer, Germany. Smith, G. 1976. Whey protein. World Review of Nutrition and Dietetics. 24:88. Stefi'e, J. F. 1992. Rheological methods in food process engineering. Freeman ' Press. In Press. Stone, H., Sider, J. L.1985. Sensory evaluation practices. Academic Press, Inc. Orlando. Stull, J. W., Taylor, R. R., Angus, R. C., Daniel, T. C. 1977. Acceptability of a whey-based quiescently frozen novelty. J. Food Protection. 40(3): 158. 132 Tamawski, V. R., J elen, P. 1986. Efl'ect of heat pretreatments on ultrafiltra- tion flux in Cottage cheese whey processing. Food Engin. and Process Applic. Vol. 2. Unit Operations :237. Watrous, G. H., Dimick, P. S., Keeney, P. G. 1991. Utilization of Cottage cheese whey. Notice of research project. American Dairy Products Institute. Weiner, G. D. 1977. Changing times for the whey industry. American Dairy Review. 39(12):42B. Whorlow, R. W. 1979. Rheological Techniques. J. Willey & Sons. New York.