Plastic Welding: Optimal Joint Strength is Rare in the Real World

by Tom Kirkland

Plastic welding is usually the last step in an entire sequence of manufacturing processes. It is also last in a sequence of design decisions that must logically begin and end with utility, marketability, and profitability. This means that, unfortunately, the possibility of obtaining laboratory-like optimal joint strength in the finished product is next to nil. That is not to say that the product will fail to meet joint strength requirements, just that optimal joint strength is seldom attainable outside of carefully controlled laboratory conditions. It is simply most often the case that less-than-optimal strength results in the best combination of utility, marketability, and profitability.

This article wasn’t written to whine about real-world obstacles to obtaining optimal weld strength. But it is important to point out the difference between the laboratory, where weld joint strength to base material strength ratios are developed, and the real world, where joint strength is one of the items on the table when compromises are made.

A search of technical papers dealing with the attainment of optimal joint strength in a variety of plastic welding processes leads to an understanding of some general rules for maximizing joint strength over all of the plastics welding processes. If you were setting up a laboratory study of optimizing joint strength of plastic parts, you would want to remove as many obstacles to optimal joint strength as you could.

Barriers to Attaining Optimal Weld Strength
There are several general areas of concern in attaining optimal weld strength. A laboratory study would seek favorable situations in at least these four.

1 Joint geometry should be consistent part-to-part and the energy path to the joint should be as simple as possible.

In the real world, seldom, if ever, are either or both of these criteria met. When near-optimal geometry does exist, it is greeted with cheers by applications engineers, manufacturing engineers, process engineers, set-up people, quality people, and anyone else who has ever struggled with a design that seems to have been created without regard for manufacturability. When such geometry is consistent across multiple mold cavities, plant shifts, resin lots, and phases of the moon, by all rights a national holiday should be declared. The plastic welding process is often where quality problems will show up, but only well after the parts are made and all hope of making more in time is gone. At that point, calling in the “plastic welding wizard” may or may not get the order out the door on time or with acceptable quality.

2 The parts should be designed purely for the purpose of welding.

This means parts should not include any pesky details like “Class A” surfaces, thin sections, curved walls, internal components fighting to get out, projections, protrusions, ribs, gates, or anything else that would require the welding settings and/or tooling to be of the less-than-optimal-for-weld-strength variety.

3 There should be an unlimited amount of cycle time available in which to obtain optimal joining results.

Study after study on virtually every plastic welding process shows that longer cycle times result in improved strength, up to a point of negligible returns. Of course, in the manufacturing world, time is money. In some cases, the shop has to make X many parts and get them out by Y, with little regard to how many parts might end up scrapped, or how close to unacceptable the welding results are.

4 Testing should be conducted in a controlled and objective manner, and 100 percent of assemblies should be tested to destruction to verify that they are, in fact, optimally welded.

Naturally, one of the more daunting tasks in real world plastic welding is identifying when a good process starts giving bad results. It also is important to recognize that criteria that is dependent on who tests the parts can create huge problems when the 98-pound inspector goes on vacation and the 280-pound inspector who can bench press 400 pounds fills in.

Case Studies

Real world manufacturing often meets with the aforementioned problems. The following case studies examine potential difficulties in more detail.

Case study number one: High-impact polystyrene is supposed to be easy to weld using ultrasonics, but a part and process designed for medium-impact polystyrene became a nightmare when first production parts started cracking after a few weeks of actual use. The solution to the cracking problem was a switch to high-impact polystyrene. The parts in question had a taper, making the joint further from horn contact at one end of the part than it was at the other. Because high-impact polystyrene does not transmit sound as well as medium-impact polystyrene, this operation became an on-going fire drill for an entire team of manufacturing engineering types who may still be searching for the solution to inconsistent welding two decades later.

Case study number two: A certain automotive engine sensor contained a coil of microscopically thin wire. The housing for this sensor was sealed ultrasonically. Approximately one in ten thousand of the sensors would fail after a few weeks of use because of a weakening of the wire by the ultrasonic amplitude during the welding process. A high-potential test would not cause the wire to fail – only the mechanical loading caused by actual engine vibrations could expose the weakness in the wire. An attempt to mitigate the problem adjusted the process to weld the parts as close to the minimum strength specification as possible. The ultimate solution to this problem may well have been the eventual obsolescence of the part.

Case study number three: A certain part having two threaded inserts could be made to over five times the minimum strip-out-torque specification if the ultrasonic insertion cycle were allowed to extend out to three seconds. Set-up people routinely set the job up with a one-second insertion time and just barely passed first- and last-part destructive strength tests. Plant management was happy with part cost (at least for the first year).

Case study number four: For a certain high-volume assembly, hot plate welding was the only practical welding method. Because of the long cycle time of the process, plate temperature was eventually turned up beyond the degradation temperature of the resin in an effort to increase heat transfer speed and reduce cycle time. The product routinely shipped with weak, contaminated joints that barely met strength specifications.

Case study number five: A polycarbonate case worked very well with a single-cavity pre-production mold, but when the multi-cavity production mold was completed, slight differences in cavity pressure profiles caused extremely minute geometry and stress variations that resulted in unacceptable “smiling” of the weld joints (a purely aesthetic issue, as strength was acceptable) that led to scrapping of millions of dollars worth of engineering time and tooling and a return to the old design that was much simpler to weld.

Case study number six: Economics dictated that wide-spec resin be used in a particular product that had a stringent weld strength requirement. Overwelding by even a little bit would cause a degraded part appearance. The variability of the resin required regular adjustment of the welding settings in a sometimes futile attempt to keep weld strength up to minimum limits. Since the only weld specification was a strength test to failure, nobody could ever say with any confidence that any particular assembly was shipped meeting the strength specification.

Case study number seven: A three-piece assembly needed to be air tight. Ultrasonic welding of two pieces trapped the third, which had an overmolded silicone gasket. Coating the gasket with silicone lubricant prior to welding increased the percentage of parts that passed the required pressure-decay leak test, but the scrap rate never got much below five percent over the life of the product. While there had originally been both a strength and a seal requirement, the strength requirement was waived in an effort to get a higher percentage of parts to pass the pressure test. The new version of the assembly, designed with weldability in mind, has run for many years with an extremely low scrap rate, and the entire assembly has much better appearance and functionality. Even without the difference in scrap rate, the new design costs less to make than the old.

So, Finally…
The point is that a good part and process design takes all competing desired factors into consideration early on. A fantastic product idea that looks great and sells like hotcakes is probably still going to be a failure on the business end if it can’t reliably be made. The best defense against all of the factors that stack up to make joint strength less than optimal is to take steps to beef up joints and create a “no doubt” assembly that can be produced reliably and consistently, and to resist the urge to over-promise on cost or cycle time. An extra nickel in material cost to get more molding consistency could be worth a dollar at the welding operation. Success in plastic welding requires that both part and process design be done with an eye to utility, marketability, and ultimately profitability. If there is insufficient balance in the compromise, the entire venture may fail.

Tom Kirkland has over 20 years of experience as both a user and supplier of plastics assembly equipment and tooling. He has been active in several professional societies, is a past President of the Ultrasonic Industry Association, and is well versed in a wide variety of plastic welding/joining processes. Tom has conducted over a thousand plastics assembly training sessions, conference presentations, and talks in many countries. He is currently a consultant in plastic welding, and is the proprietor of, a supplier of parts and supplies for plastic welding.