5-8 ENGINEERING PROPERTIES OF LAMINATES 



The flexural strength data for woven roving showed an important difference with 

 laminate thickness. Flexural strength data for laminates with other reinforcements, 

 2 ounce mat and 10 ounce cloth did not vary with thickness. Thickness effect for woven 

 roving and mat laminates faced with 10 ounce cloth was the same for equivalent lami- 

 nates without the facings. 



The flexural modulus values were affected by laminate thickness for all reinforce- 

 ments and angles to the warp. 



Compressive strength was found to vary with each thickness for all reinforcements 

 except 10 ounce cloth. 



For shear strength perpendicular to warp, 10 ounce cloth and woven roving lami- 

 nates showed some differences with thickness while mat laminates did not. 



For shear strength parallel to warp, laminates with all types of reinforcements 

 showed small differences with thickness. 



Discussion of Tables 



The data are presented in a manner considered important, from a statistical viewpoint, 

 to clearly indicate the effect of different shop practices on the properties of laminates 

 molded by the contact or hand lay up method. For most properties, the tables contain high 

 and low values as established by the statistical analysis. An average value and a lower 

 limit value established at a 95 per cent confidence level is given for each range. It is 

 important to emphasize that both the high and low range of values were determined from 

 test panels made by good fabrication practice as individually established by the participating 

 fabricators. The high range of values in the tables may be taken as high standards for good 

 fabrication practice and experience with the particular type of reinforcement. 



The data in the tables would present a much simpler appearance if single property 

 values were given for each material. Due to the wide differences in the properties of the 

 laminates produced by the participating fabricators, the use of a single property value 

 would not be valid and could lead to serious misunderstanding. If the average property 

 values of the laminates in the lowest range were reported alone, the tables would give no 

 indication of what can be accomplished by good shop practice. If the average value from the 

 high range of properties was used as a single value, designers would be misled into thinking 

 that all fabricators could produce laminates of equally high quality. If an over -all average 

 of the high and low test results were reported alone, it would differ considerably from the 

 true values of any fabricator. Any attempt to rectify this situation by establishing some kind 

 of a range of variation around an over -all average would introduce more confusion, since 

 this range does not represent the variability of the material of any one fabricator. Rather, 

 it would be a mixture of the intrinsic and random variability of the material for any one 

 fabricator with the systematic differences in level due to systematic differences in 

 shop practices. 



There appears to be no valid substitute then for indicating the variation among fabri- 

 cators, particularly when this variation is large. Therefore this variation is indicated by 

 giving two separate average values. One value is for those fabricators who maintained a 

 definitely higher average level; another value is for those fabricators who produce material 

 at a lower average level. 



