Tobacco Documents Online

0 '::" E 0 ,n ~."6 t/) "C "u"~ (1) ~-" "='E "E: :D L~J E' E [ 'E o :'El -J ~D rY I m OOO~OOOOOOOOO (%) a3uaJa~ip #nd Jad ~aA!lap a^! ~laJ u3 t~ O g ~g O ~D O c~ O U3 u~ G) O O -[ l cn I E O E 0) o O O t) O O l-- E <, ctl 5S2301435

13- C d <: C .q ,_ E O-13_ o>E c "6:a ~o (h e) O {::: O ~J > ~3 O ,o E] | Q K E :jLJ G) 03 0) C C O O E C Illll I (%) 8-3ueJaJ~!p L~nd Jed ~JaA!lap aA!ISIaJ ~L e~ o .°E ~E e£ ~E 8> o t~ ul u~ r~ 0 I E 0 E O3 0 0 0 0 O) E <, ca 559301436

BAT GROUP TECHNICAL PROGRAM PROJECT BRIEF DATE: 07/07/95 Work Area: 01 Product Technology Key Activity: 01.07 Cigarette Design Group Coot.: W. Sclmeider Project Number: 01.07.04 Location: BAT-G Title: Design-Sensory Atlributes/Models Proj. Leader(s): W. Schneider Local Proj. Designation: PV 03.003 O biective/Business Relevant: o translate smokers' description into useful data that can be utilized by the product developer. models which assist to develop and modify products more specific to corxsuml feedback. To translate smokers' description into physical/chemical clgarerte and smoke data a model is used which eontalns the dependences of smoke deliveries on cigarette design and allows to consider smokers' larger values as well as upper and/or lower limits of zones of acceptance for several "restricting parameters". The currently used auxiliary criterion for the target value is condensate per puff, and the currently used reslrlcting parameters are puff volume (upper limit), pressure drop (upper limit) and nicotine per puff (upper and/or lower lirnlt). Measurements of smoke deliveries (beyond condensate and nicotine) are performed (in other local projects) to provlde Ibe chance to identify further (hel~er) target and restricting parameters (eg. indicators for irritation). The required experimental input information on tasle dimensions is derived from panel results, The model is used to check or to establish hypotheses for current taste assessments and their assoclatlon with design features. The model is applied for qualitative (not quantltative) evaluation. Two types of evaluations are per fornled: (1) The ranking of brands according to a specific taste dimension is compared with different ankings uf these brands accnrding to their smnke deliveries, which result from different empirical assumptions on target and restricting, values. Those target values and restricting values are supported which provide the best agreement betxveen ranking of taste dimension and smoke debveries. This may open an oppo~unity to identlfy classes of smokers (by their target and restric(h~g values). (2) Applying target and restricting values from best current kanwledge Ibe position of Ibe brands retative to Ihe upper and/or lower limits of the znnes of acceptance can be determined. This can be compared with the aceeptabilty of the brands (assessed by panels). This t~pens an t~pportunity to design cigarettes in these zones of acceptance which are identified for a ce~aln class of smokers. These evaluath~ns have been performed for low tar cigarettes. Collecting and assessing further experitnenlal data from all sources, which contribute to valblate cr to modify the model approach. Application to current ultra low tar brands. ~- ,#, 1437

6'1 ¢4 C) Methods for evaluation of complex interrelated data: cigarette and smoke attributes .H- sensory attributes advantages I disadvantages: • statistics / multivariate analysis • neuronal networks • knowledge based systems ("rules", "fuzzy knowledge", "qualitative simulation",...) British-American Tobacco (Germany) GmbH 1995~

-2 - ~ontent i. Meaning Objectives Special features of the CONSidER DEMAND RESPONSIVENESS approach 3.1 Link of consumer descriptions and cigarette data 3.2 "cause ~nd effect" relationships 3.3 Additional informations 3.4 Different smoking parameters 3.5 Different types of smoking behaviour Quality of target delivery 3.7 Quantity of target delivery Restrictive properties Expressions related to CONSUMER DEMAND RESPONSIVENESS Areas of applications / Examples 51 Identification of potential product advanrages 52 Explanation of and feed back from consumer assessments Feed back of information from successful brands 5.4 Brand dsvelopl~ent Appendix A: Definition of "elasticity" Appendix 5: Supporting tools 55 01439

r%~p° criterion only the condensate delivery per puff i~ used as target delivery, Curren~ investigations will provld~ mQre specific chemical compllment~ry infozmations (which can be £reated by the f0r~alism in the sa~e way as the condensate ~el~very per puff}. The comparing of two brands, based on the condensate dsli- veries, can only be mads, if the types of co~densates Qf the twe brands are not too different I~-e* the blends of tbH £wo brands must be similar). ~f ta ~ delva • The quantitative aspect is described by the equation M(i,V}/pn(v) = M(i,vo)/pn(vO) + E * (V - v0) {!) M(i,v): delivery of s~oke ~onstituen~ i per ~igaret~e at puff volume V ~n(v): puff number at ~uf~ vol~me v V: actual puff volume (ml) V0: standard puff volume ( 35 ml} ~: ~ean el~st~city ~f the cigarette (definition see appendix) In this ~ategory an average on the deliveries o~ all puffs of a cigarette is considered. In extended ~pproaches deliveries of 1'early~, ~nd "late" puffs are di~crimlnated. 3.B ~Bstrlc£ive ro ~ ti~s the in~rease of the p~ff volume can be restrlcted by s~v~ral ~roperti~s~ (ii Th~ ~ff vol~me itself is a ~estrictive prope~y b~ca~e of ~he physi~loqical co~diti~n~ o~ ~h~ s~ok~r. 552301440

Example for application: Evaluation of panel results of ULT-brands O~ectives: (I) Comparing ranking of brands with regard to (1) fullness (2) strength with ranking of brands with regard to condensate and nicotine deliveries per puff Can the rankings with regard to fullness and strength be explained - by perpuffdeliveries at standard conditions? if not: - by per puff deliveries at other fixed puff volumes? if not." - by target deliveries and/or restrictions? - by which target deliveries and/or restrict/ons? British-American Tobacco {Germany) GmbH

II) Identification of those brands the ranking Of which cannot be explained by condensate and nicotine deliveries per puff (even if target deliveries and restrictions are considered) [111) Approach to explain the ranking by different smoke quality - total tar headspace per condensate L [~r~-)_ British-American Tobacco (Germany)GmbH --1995 -- 5t~Z30144Z

IFullness versus condensate per puff (ISO)i 5 I,o c 4,5 4 ~M~dL 3,5 3 PMLO R10 LUO I I I 0,1 0,2 0,3 0,4 0,5 NFDPM(mg) ~-- Tobacco (Germany).GmbH British-American 0,6 0,7 1995

c- O3 r- O3 2,6 strength versus nicotine per puff (ISO) I R1 © 2,4 2,2 2 1,8 1,6 1,4 0 PM~ PMU© ~er © O LU AdL~ 0,01 0,02 0,03 nicotine (mg) 0,06 4~ 4~ British-American Tobacco (Germany) GmbH 1995..,-...~