MIL-HDBK-17-1F Volume 1,Chapter 3 Evaluation of Reinforcement Fibers method as applied to fiber,a test procedure has been developed within the MIL-HDBK-17 Testing Work- ing Group (see Section 6.4.4.4.1). ASTM Test Method D 3800(Reference 3.3.2.1(e))deals specifically with obtaining the density of fi- bers.This standard covers three different liquid displacement procedures:Procedure A,which is very similar to the D 792 liquid displacement method (Reference 3.3.2.1(f);Procedure B,in which a low- density liquid is slowly mixed with a high-density liquid(containing the fibers)until the fibers become sus- pended;and Procedure C,which simply references D 1505,which is a density-gradient method. For detailed guidance on D 1505 and helium pycnometry,the reader is referred to Sections 6.4.4.3 through 6.4.4.5 of this volume of the Handbook.Note that Section 6.4.4 refers specifically to composites, but the methods discussed are fully applicable to fiber measurement except as noted below in Sections 3.3.2.2 through3.3.2.3. 3.3.2.2 ASTM D 3800,Standard Test Method for Density of High-Modulus Fibers The approach taken in ASTM D 3800 is threefold.Procedure A is identical to D 792 except that the immersion fluids recommended have only fibers in mind.The concern is complete fiber wetting and avoiding entrapped microbubbles.Procedure B relies on careful mixing of two liquids of different densi- ties (with the fiber immersed).When the fibers are suspended in the mixed liquid a hydrometer or liquid pycnometer is used to determine the density of the liquid.The density of the suspended fiber is equal to that of the liquid.Procedure C is D 1505 inserted as a part of D 3800 by reference. Given that apparatus and procedures are identical to D 792 for the liquid displacement procedure (Procedure A),and that Procedures B and C have much in common with D 1505,the reader is referred to Sections 6.4.4.2 through 6.4.4.5.Here,only those test aspects peculiar to fibers are discussed. The experimenter needs to be mindful to avoid entrapped bubbles,liquid absorption,and problems involving the fiber sizing coating (if any).Common sense immediately flags roving as a difficult fiber form to wet out,yet complete wetout is required to produce meaningful data.Pay close attention to the inter- filament regions.In D 1505 the problem is not as severe because the fibers can be cut and/or spread out prior to insertion.Since the measurement is direct the size of the fiber sample is irrelevant.Immersing many small fiber fragments allows for direct verification of density variations of the fiber,keeping in mind that small fragments may take hours to sink to their equilibrium density level.It can not be emphasized enough that complete wetout must be achieved.Use of high-wetting,vacuum-degassed liquids go a long way to this end.Remember that the fibers are a prime geometry for nucleation of gas bubbles out of so- lution.If the liquid is not fully degassed a bubble-free roving can quickly form new bubbles. The surface area to volume ratio of composite fibers is extremely high.For cylindrical shapes, S.A./V=2/R,where R,the radius,is only several microns.For a 0.028 mil (7 micron)fiber the ratio is 143,000 to 1.It is,therefore,very important to ensure compatibility between the fiber and liquid.Glass and polyethylene fibers are fairly immune in this regard;however,aramid,for example,is certainly not. The liquid immersion time should be kept to a minimum to avoid liquid diffusion into the fiber. The mistake is often made of thinking of the fiber by itself,when in reality it is usually coated with an interfacial sizing agent(to promote improved bonding with the matrix resin).It is good practice to re- search the sizing agent,as it is a completely different material than the fiber(with different absorption and chemical characteristics).Since the sizing is applied to the outer surface of the fiber even the volume of a thin coat quickly becomes significant.For example,a 0.028 mil(7 micron)diameter carbon fiber with a typical coating of 1%sizing agent on a weight basis (with assumed density of 1.2 g/cm(0.043 Ib/in ) gives a final product which is 98.5%fiber and 1.5%sizing on a volume basis.For precision work,strip the sizing agent off the fiber before measuring fiber density. 3-7MIL-HDBK-17-1F Volume 1, Chapter 3 Evaluation of Reinforcement Fibers 3-7 method as applied to fiber, a test procedure has been developed within the MIL-HDBK-17 Testing Work￾ing Group (see Section 6.4.4.4.1). ASTM Test Method D 3800 (Reference 3.3.2.1(e)) deals specifically with obtaining the density of fi￾bers. This standard covers three different liquid displacement procedures: Procedure A, which is very similar to the D 792 liquid displacement method (Reference 3.3.2.1(f)); Procedure B, in which a low￾density liquid is slowly mixed with a high-density liquid (containing the fibers) until the fibers become sus￾pended; and Procedure C, which simply references D 1505, which is a density-gradient method. For detailed guidance on D 1505 and helium pycnometry, the reader is referred to Sections 6.4.4.3 through 6.4.4.5 of this volume of the Handbook. Note that Section 6.4.4 refers specifically to composites, but the methods discussed are fully applicable to fiber measurement except as noted below in Sections 3.3.2.2 through 3.3.2.3. 3.3.2.2 ASTM D 3800, Standard Test Method for Density of High-Modulus Fibers The approach taken in ASTM D 3800 is threefold. Procedure A is identical to D 792 except that the immersion fluids recommended have only fibers in mind. The concern is complete fiber wetting and avoiding entrapped microbubbles. Procedure B relies on careful mixing of two liquids of different densi￾ties (with the fiber immersed). When the fibers are suspended in the mixed liquid a hydrometer or liquid pycnometer is used to determine the density of the liquid. The density of the suspended fiber is equal to that of the liquid. Procedure C is D 1505 inserted as a part of D 3800 by reference. Given that apparatus and procedures are identical to D 792 for the liquid displacement procedure (Procedure A), and that Procedures B and C have much in common with D 1505, the reader is referred to Sections 6.4.4.2 through 6.4.4.5. Here, only those test aspects peculiar to fibers are discussed. The experimenter needs to be mindful to avoid entrapped bubbles, liquid absorption, and problems involving the fiber sizing coating (if any). Common sense immediately flags roving as a difficult fiber form to wet out, yet complete wetout is required to produce meaningful data. Pay close attention to the inter￾filament regions. In D 1505 the problem is not as severe because the fibers can be cut and/or spread out prior to insertion. Since the measurement is direct the size of the fiber sample is irrelevant. Immersing many small fiber fragments allows for direct verification of density variations of the fiber, keeping in mind that small fragments may take hours to sink to their equilibrium density level. It can not be emphasized enough that complete wetout must be achieved. Use of high-wetting, vacuum-degassed liquids go a long way to this end. Remember that the fibers are a prime geometry for nucleation of gas bubbles out of so￾lution. If the liquid is not fully degassed a bubble-free roving can quickly form new bubbles. The surface area to volume ratio of composite fibers is extremely high. For cylindrical shapes, S.A./V=2/R, where R, the radius, is only several microns. For a 0.028 mil (7 micron) fiber the ratio is 143,000 to 1. It is, therefore, very important to ensure compatibility between the fiber and liquid. Glass and polyethylene fibers are fairly immune in this regard; however, aramid, for example, is certainly not. The liquid immersion time should be kept to a minimum to avoid liquid diffusion into the fiber. The mistake is often made of thinking of the fiber by itself, when in reality it is usually coated with an interfacial sizing agent (to promote improved bonding with the matrix resin). It is good practice to re￾search the sizing agent, as it is a completely different material than the fiber (with different absorption and chemical characteristics). Since the sizing is applied to the outer surface of the fiber even the volume of a thin coat quickly becomes significant. For example, a 0.028 mil (7 micron) diameter carbon fiber with a typical coating of 1% sizing agent on a weight basis (with assumed density of 1.2 g/cm3 (0.043 lb/in3 )) gives a final product which is 98.5% fiber and 1.5% sizing on a volume basis. For precision work, strip the sizing agent off the fiber before measuring fiber density