CASE STUDIES: Solvent Cracking

SOLVENT CRACKING IN CPVC TANKS
A long-time PELabs client — a custom fabricator of built-up plastic structures — discovered cracks around threaded inserts in flanges of a recently fabricated CPVC tank. The cracks were radial, emanating from the circumference of the hole for the threaded insert. The cracks didn't appear immediately, but after a few weeks, during which time the tank was pressure tested. Since the cracking was clearly related to the hoop stress at the hole, one would be suspicious of the interference at the hole-insert interface; but the testimony of the shop personnel was that the inserts were loose in the holes, so loose, in fact, that the assemblers used a liberal amount of their standard adhesive (Adhesive A) to lock them in their respective holes.

Figure 1: Cracks Around Bolt Holes in the Flange of a Fabricated CPVC Tank

PELabs engineers immediately laid out a testing and investigative program aimed at uncovering the cause of the cracking. They included issues related to design, materials, processing and environment. The client supplied drawings, residual CPVC sheet from the tank-production lot, newly purchased bolts and inserts, and a sample of the "standard" adhesive (Adhesive A). Examination of the threaded inserts as supplied by their manufacturer revealed the presence of an adhesive (Adhesive B) on the threads to help secure the insert in place.

MATERIAL IDENTIFICATION & PROPERTY TESTING ASTM Type I tensile bars were die-cut out of 1/8" plaques machined from thicker CPVC sheets taken from the client's residual stock. Next, Izod bars were machined out of the plaques for use in impact testing and stress-crack testing. Samples of the CPVC were also sent to PELabs' chemist for Infrared Spectroscopic (IR) identification and ash analysis. Samples of Adhesive A supplied by PELabs' client and Adhesive B scraped from the threaded inserts were also sent for chemical constituency analysis.

The IR trace of the fabricator's CPVC sheet was compared with a standard CPVC and found to deviate only in a slight carbonyl shift, indicating some (probably thermal) processing degradation. The ash content was 3.65% by weight, less than the standard at 4.20%. The inorganic matter in the sheet material was found to be Gypsum and TiO2 (Talc and TiO2 in the standard resin). Adhesive A turned out to be cyanoacrylate, while Adhesive B was an oily ester mixed with an inorganic carrier. The oily ester is known to be a stress-cracking agent for many polymers.

PELabs technicians measured density, notched Izod, and tensile properties of the CPVC sheet from the client's shop. When our engineers compared these results with the material supplier's data sheet, the laboratory data were within 5% of the corresponding data sheet values. This rendered processing or environmental degradation unlikely sources of the problem. The cracking, therefore, did not appear to be material related.

STRESS-CRACK TESTING
Table I shows the results of PELabs' stress-crack testing. Ten of the Izod bars that were machined from the plaques of the client's CPVC sheet were bent to 0.9% strain in the fixture shown in the photo of Figure 2 . Adhesive A was applied liberally to four specimens, Adhesive B to four, and nothing to two specimens used as controls. The adhesive compounds were applied at the place of maximum tension on the specimens. Observations concerning cracking are recorded in the Table for various exposure times. A blank cell means no cracking was observed at that time for that specimen.

All the Adhesive B specimens cracked within 24 hrs. The cracks grew and multiplied with exposure time. Even specimen Adhesive A-1, which was contaminated with a small amount of Adhesive B, cracked, although the cracked region hasn't grown since the contaminant was removed. The Adhesive A does not appear to be a stress cracking agent for CPVC. The condition of the samples after 1000 hours of exposure is shown in Figure2.

Figure 2. PELabs Stress Cracking Fixture with Samples after 1000 Hours

STRESS CRACKING AROUND THE HOLES
The severity of stress cracking is influenced by the concentration of the chemical agent and the applied stress. Thus, in the case at hand, the fit of the inserts in the holes, and subsequent loading of the bolts in tension, increase the circumferential stresses that, combined with the chemical agent, propagate cracks. PELabs technicians therefore cut two 3/4" thick plates from the client's sheet stock, drilled and tapped (according to the drawing specification) ten holes in each plate, and measured the inside diameters of seven. Next, they measured the root diameters of each of five threaded inserts.

The results indicate that there should be no interference between the inserts and the threads, but clearance instead. In point of fact, however, with the Adhesive B still on the threads, four of ten inserts took greater than 75 in-# to run down. With the Adhesive B removed, the inserts could be run down with the hand.

CONCLUSIONS The source of the cracking in the CPVC tank was evidently the adhesive supplied with the threaded inserts. That compound is an aggressive stress-cracking agent for CPVC. The cracking may well have been exacerbated by the pressurization testing, since pressure applies tension to the bolts, which increases the hoop tension around the hole.

PELabs' client has constructed numerous CPVC tanks without Adhesive—B since. All passed the pressurization testing and are in service without any evidence of cracking.

markets served  ·   services  ·   case studies  ·   company profile
contact us  ·   tech resources  ·   glossary  ·   did you know  ·   what's new  ·   product development

© 2003 Plastics Engineering Laboratories.
PELabs respects your privacy.
We will not share your information with any third party. Period.
Powered by media firma.