Advanced Composites Analysis Program (ACAP)
The ACAP finite element program was developed as a proof-of-principle to demonstrate the integration of a unified hygrothermal (i.e., combined moisture plus thermal) effects model for generally orthotropic composite laminated materials. It combines several smaller algorithms readily available on this website with an isoparametric finite element derivation for a hexahedral (solid) element having a basic eight-node design. Extension of the program to examine a 'serendipity' element (a basic hexahedral with edge mid-nodes for a total 20 nodes) as well as a quadratic 'Lagrange' element (the previous hexahedral with all mid nodes for a total 27 nodes) is proposed for future development - time permitting.
The proof in the ACAP model was to develop not only the mathematical derivation of a generally orthotropic hygrothermal finite element, but the supporting framework capable of analyzing a variety of real world problems, such as the sophisticated (but not too terribly difficult to model) composite missile interceptor cone shown in the figure above. The impetus for the program development was actually, a sort of dare, brought about through our discussions in 2009 with the U.S. Army Future Combat Systems Brigade Combat Team.
Back in that year, ISE bid on a Small Business Innovation Research (SBIR) project with the above agency (SBIR A09-197). The Army's request under this proposal was a little unusual from the start in that it was written in such a way that unless the solicitor worked for one of two different commercial finite element companies (the admitted favorite finite element vendors of this particular sector of the Army), one would have a pretty thin chance of winning the contract (for starters, you would have to have carte blanche access to these two companies proprietary computer source codes to figure out what had already been done - huh?). Undaunted, ISE proposed a course of research for a stand-alone computer code, namely, the ACAP program. ISE has done extensive work in the fields of finite elements, composite structural analysis, and particularly the environmental (hygrothermal) effects of composites, so this was more of a 'bolting the pieces together' project for us than anything else, and we already had most of the pieces in front of us. Needless to say we didn't win {surprised?} and the researcher we spoke to on the debrief actually refused to tell us which two companies he awarded the work to (hmmmm....). He concluded the call by saying that no one could write a finite element code for $800K (the ultimate development contract value through all future phases and also demonstrating his complete lack of judgement when it came to spending taxpayer dollars). We decided to prove him wrong and did the whole thing on our own nickel for orders of magnitude less - sorry, but we told you so...
The model run shown is for a hollow 3D composite missile cone constructed out of MMS-545 Graphite Epoxy Pre-Preg (symmetric +/-45 alternating laminate plies) and having four externally wound composite tanks held in place with two stainless-steel straps on a central support core of the same material (an overly simple, but accurate representation of the real thing). The composite laminate was exposed to a 50% moisture content and uniform 200 degrees Fahrenheit. The color scale is numerically arbitrary (for graphical effect), but demonstrates the relative rate of thermal evaporation of the cone at one instant in time (specifically, one hour after removal of the external heat and a slow return to room temperature). The speckled pattern is due to the uneven drying of the material and is attributed to the incorporation of a random surface roughness/porosity coefficient in the model (i.e., in real life, no composite material is perfectly smooth and this simple adaptation allows for graphical representations more closely resembling what one would see in a laboratory setting). Of course, because we're dealing with a finite element formulation, full element displacements, strains, and stresses are known for each node as a function of the drying time. Not too bad for a first attempt, and completely consistent with theory and observed test samples.
Current version is 1.0 (beta). Ongoing work as applications present themselves.
Platform: PowerBasic CC for Windows XP.
