Document RpLom5Qpqd0DJ7jQBRv2L6dkX

i V? ~>Mr. F. A. Barker Mr . Jack Clegorn January 21, 1947 rjdf'iV.. . . xw? ^1 t* . vV- -- ;Mr. J,- Fred Beeves GBHEBAL ELECTRIC - PETSFISID c? *1#? *i'f-C*Kv* ' "" asIa referenc# to.Hr; tpieelo<*if'letter of January 15, Attached, 1 have dieeussed the Annistoni analysis ofAroclare VithMr^w ?4 V5 ,, Cfl, i*.v#fL . %*$ /ii-p. |U ~~^"*jrS2 &*-?* fs. Clegorn, Sir* Jenkins and Mr, Barker and I do not Believe that Anniston is far off Base, If at all, on our Aroolcrs or their analyses* Ton will note Mr. Barker's letter, attached, concern- lng the pour point of our Aroclor 1254. Mr. Barker has checked pour points on the same material By five different people who formerly analysed drooler regularly, These five people Inde pendently arrived at a 10q pour point on the same sample. ~ Our laboratory, using the ASTM method and ASTM standards, states that Individual results on the same oil In any one laboratory may vary by 5F, whloh is 2.8^3 and that different labcratarles may vary By 109F, which la 5,6C although the average of three or more results in different laboratories should show the differences Between averages no greater than > which is 2.8C. J ... The test that Mr, Saab ran shows a reproductivity of results of -4 plus or minus cflC and you will note that the tabulation of our pour point with the G.S. pour points varies By only 1.89F on the average. This is well within the limits allowed Bj dSTK standards. . . - y-V : .! m/' . :. * Yi; The ASTM standard method calls for the use of ice to cool the sample. In the range for Aroclor 1254 the G.S. analytical calls far the use of liquid air or liquid oxygen. It Is my belief that the cooling By liquid, air Is very rapid at the outside of the tube and that the Aroclor can Be stiffened Before a true and correct temperature is shown on the thermometer near the center of the tube. The use of Ice does not allow such rapid chilling and I Believe that our thermometer will probably read closer to the true tempera ture of the Aroclor. In any case, If G.S. uses liquid air and we ' latC7, V *; vV<V^F/- ' use Ice, this. Is probably the .reason for the difference In pour .. v pointsi A Furthermore, I d^|o^|k|xcwiwhdt3^?r.'|t''.E. haA^s|ted their^|;|j; -?V 'll,. method wlreproductivity .ofiresults as/haaBu* labaw^ary*: have not 1 Believe that this should Be done, - vi` they./'. .. * ^V4** , i-S'-ts* jvj G.E. also states that Anniston resistivities of 300 z 10^ Is Below , specification. This is not so as our specification statea that resistivities shall Be above 100 x 10^. I am sure that Both G.S. and our Sales Department Is aware of this. I am unable to see why there Is a difference between the resistivity results of our labora tory and G.S. They state that we may have a dirty cell. It may DS\N 330987 STLCOPCB4077631 T i Mr. J. I. Beeves * Pags #2. 1/21/47 also 1>e that there 1b a difference In methods or In the machines used for this test. Z have been unable to find anything about this. : - ; r* . V \ ' ' ` - - " rikyS r M *4*h General Jleotrlc has. also stated that a value of .,125 for /Chlorides, * *> t **l 4- 1'*' *4** 1 * - /2//<~v - t ,. _ 'u'^' :^_^ >. . - -s. ^ in the corrosion test,Is above specification, Shis Is not so as : , : * * they *do?" fojt&i^'^tes^^^^|^^aat1^:|^^dlse^i^';a^ great! length some >ears ago and that' it was discovered that due to differences in the quality of vlcoelmeter or to the differences in methods It was impossible to check on this determination, I do not know the ..... whole story but Hr. Jenkins has it in his files and has promised to , send it to me. However* It seems that Q.Tt, is aware that they normal ly get approximately 15 seconds higher and Z do not understand why G.l. now complains about a difference. / 1 , .. Z note that on the third and last page in the table of defects that the refractive index 1b above specification for two lots. The speci fications for refractive index for Fyranol 1467 are 1.6137 to 1.6l47, Z do not understand why these lots are high. , GJC. evidently Is worried about a discrepancy in analytical results rather than about any defective or off grade Aroelor or they would be complaining about the quality of the material rather than about analytical results. With the exception of the refractive Index of 1467 and poesibly the resistivity of 1254, I see no serious error . on the part of the Anniston plant or laboratory and it Is quite like ly that a careful cheok of resistivity will reveal a condition similar to the pour point and viscosity determination where the differences can easily be explained. - /* * ,, -. < k, ,v yiy* fsii ' %. "* y ,,< ^ . '/ 1 V .*% > , v,lt. ?. Leaven #gtT' ^ fl r * ____ - . DSW 330988 STLCOPCB4077632