Document 3QK09go9pZEbZNz1VX68ddRQE

W-' QQt J, p, HoBarmy:} $11. P~7t fi\t.0 loW.&rg^hV*nijt,p LMaS4bV .> E. 3. Davis, lab. f. B. Evans, Lab. Ifl IBSBALl LiBfliTiy (JBBJWf 0. E E. Merklo, Lab. R. E. Bike, Lab. 1. P. Morse, Lab. B* L* Selby, Lab. Indexed E Copy . w an asmAim- Marshall Laboratory larch 21, 1952 BBHORAHDlgg REPORT NO. 1-439 "s t r e s s o r kCKm" o p ms ix t i^e c t x &h b s hAVy.v *I, /*' i>27 ?5<} XMODtJOTlOMt There? have been a number of complaints recently on the cracking air-dry of Type H "puco" furniture finishes resulting fro & reformulation to use soya beau oil plasticiser instead of the castor oil which was in short supply. This work was undertaken to investigate the relationship between this field experience and certain physical property tests to determine the major factors which influence this cracking. SMi&RY hm COMPLIESXOMS i 1, Air dried soya bean oil modified furniture finishes appear to have an exaggerated tendency to crack when subjected to thermal or mechanical shock. This is attributed to unrelieved film stresses and therefore the ter "Stress Cracking;" is proposed-to describe the resulting phenomena. 2, This cracking tendency may be greatly reduced by short period heat aging. The reason for this is not clearXy understood, but apparently some incompatibility of nitrocellulose and plasticiser is corrected by heating. The use of the term "annealing" for this process is proposed. 3, Preliminary data indicates that "Stress- Cracking" , tendency may best b measured fey a graph of the maximum film build which will pass a on cycle cold box test vs. the age of the film when tested. A tentative test method is attached to in this report, 4... The conventional "cold crack." test measures the cracking tendency of an annealed finish when subjected to . alternate hot and cold cycles * A form of fatigue failure is measured by the number of- cycles necessary to cause failure.:..,. These preliminary data indicate that'"stress crack" resistance y bo mors significant than conventional cold crack resistance in predicting the field performance of a furniture -finish. 5. Previous workers have found that the "K&llen" cold distangibility tests give a satisfactory correlation with field experience on the "cold crack" resistance of annealed, finishes. The current difficulties do not invalidate this compilation. N35444 6. Preliminary data indicate that the "Stress Cracking" resistance of both the castor oil and soya bean oil products may be improved by ''annealing", but the annealed soya formula remains inferior to the air dried castor oil formula, 7* She experimental work described below is only of a "high spot" nature. It is not intended to serve as an evaluation of any teat method but was dotis to develop leads which might point to the methods worth evaluating, EXPSR XMBHTAL: Test k. listensibilltles were determined on copper panels as deioHBed by R W. LaBerge in Research Report #1142 and TP-51.2B-51 except that the 14200 "d u c o " was applied by spraying rather than by the pour down method. The topcoat was applied directly to the copper. Panels to be baked were flashed one hour at room temperature then placed in the Haas oven for X~X/2 hours at 130*F and 35^ KH, Other panels were kept at room temperatures, After aging, the panels were placed in the Mullen Tester at ~20F. for 1-1/2 hours before the test was conducted. Most of the panels cracked before the test was run. These account for the blank places on the data sheet. The age given on the data sheet is the time between spraying and placing in the Mullen Tester. Test B, Cold tests and mechanical flexing tests were run on p&Wbgany panels. Large 15 x 24 inch panels were prepared with one formula on each half The panels were stained with ?S"4489j > wash coated with 1366 "Butaeite" sealer, and filled with 27-5017 filler. Half of the panel was then sealed with 2 mils f the soya been oil type of 1.9380 sealer while the other half was sealed with the castor oil formula. Topcoat fils of varying thicknesses .of the two 14200 "Due" formulae were then built up on top of their respective sealers. The build given on the data sheet refers to the topcoat build, neither the sealer nor the topcoat ns sanded. The topcoat was applied in double passes with 15 minute flash periods between each pair of passes, - The panels were cut into 2 x 12" pieces for testing. Duplicate cold test panels were run in each case. The panels were aged for the period of time specified on the date sheet, ' then subjected, to a one cycle cold test. Sixteen hours at ~X2*F was used for the 46 and 112 hour old panels while six hour at -19 F was used for the 16 hour old panels (The~12P cold box was not available). The panels were then brought to room tempera ture, All cracking occurred with 1-1/2 hour after removal fro cold box,. Test 0, Preliminary work ted shown that cracking of the type Observed in the field could be produced fey bending a 2 x 12" pane2 over the corner of the laboratory bench without using any refrigeration equipment. This procedure was followed on a DUP030003367 y' 3 set of panels identical to the panels suh^octefi to cold crack tests In general the results of this flexing were in fair agreement with the cold crack tests. It is apparently a crude test which can he used to determine gross differences in cracking resistance. FESstub DUP030003368 4- CRACKING RESISTANCE OF SYSTEM 19380 SEALER U2Q0 TOPCOAT 0 m cracked NO m m cracks AGS as exact age of cold dlstensiblllty panels and approximate age of other panels Duplicate wooden panels run for cold crack, AH duplicates agreed except in the three cases noted. Range of 95$ reliability of distensions is approximately 4; ,003 Inches, DUP030003369