by Tom Kirkland, Tributek
The first experience many of us had working with disassembling and assembling stack components for ultrasonic welders was when tooling needed to be changed in the shop. We were led to the machine by the company’s ultrasonic guru, handed a set of hex keys and some funny-looking spanner wrenches and told to start taking things apart.
Once the stack was out of the machine and on the bench, we first had to puzzle out which way to put the spanner wrenches on. I still don’t understand how any mere mortal could get these pieces of tooling apart without acquiring broken fingers, a hernia or an aneurysm. We then did whatever the current guru recommended when we put the tools back together, and we assumed from that point forward that was the way to do it.
All was well and good until the first time we encountered somebody – maybe an ultrasonic rep or someone we met at an SPE meeting or tradeshow – who loved a method other than the one we had been taught. The frustrating thing was that there seemed to be no way of knowing who was right.
The fact is that there are several schools of thought when it comes to ultrasonic stack joint assembly, and often more than one approach can deliver good results.
All ultrasonic stack joints are threaded joints. Two flat mating surfaces are brought into intimate contact by a single threaded connection in the center of the interface. In the case of a replaceable tip, the threaded connection is actually made in one piece with the tip. In the case of a horn/booster or booster/converter joint, the threaded connection is made with a separate threaded stud and both components have threaded holes to receive the stud.
The flat mating surfaces must be in such intimate contact that the high power acoustic energy is transmitted from one component to the other as though the joint did not exist at all – in other words, as though the two components were made from a single piece of metal. Even the smallest relative motion between the two components will result in problems including, but not limited to, the generation of heat. Generation of heat in an ultrasonic stack is almost always bad.
Replaceable tips are used on smaller titanium horns to allow for changing out of the horn face when wear to the face occurs or when a different detail is needed. With recent advances in tool steel horn technology, the use of replaceable tips has been almost completely banished from high-volume production and left to low-volume production and lab use. This is a good thing, because replaceable tips can cause significant problems if they are allowed to get hot, which can occur quite easily in volume production. Replaceable tips are treated somewhat differently than other stack joints and fall a little outside the scope of this article.
The reader may have noted that no matter how careful one is to not over-torque the joints when putting them together, supreme effort almost always is required to get them apart. The issue in many plants seems to be how to get the joints apart without damaging the components. Tooling vises appropriately sized for ultrasonic tooling are widely available, and they work well and are highly recommended. Also, use of the tools recommended by or manufactured by the company that made the ultrasonic machine or tooling will generally give best results. Pounding on wrenches with mallets or hammers is to be considered a last resort. Do not let the amount of effort some joints take to disassemble sucker you in to using too much torque on reassembly.
Some people recommend titanium studs and others recommend steel. The advantage of titanium studs is lower mass. Especially at high amplitudes, lower mass in the stud can be a very good thing. The tradeoff is that titanium, while high in fatigue strength, has considerably lower tensile strength than steel, and tensile strength is more important in this case. So a titanium stud generally has to be larger in diameter than a steel stud for a given application, which takes surface area away from the flat contact surfaces that will transfer most of the energy. Titanium studs are made of a much more expensive material and must always have cut threads, never rolled, so they are considerably more expensive as well. The opinion of this author is that there is not much, if any, advantage to be gained from the more costly titanium studs.
Before considering how the job is to be done, make sure the stud holes are in good condition, without broken, missing or replacement threads, etc. Good results depend on getting all conditions as close to optimal as possible. In other words, doing it right is far preferred to “getting by with” something.
There are two schools of thought on studs – the free-floating school and the tightened-to-the-downwind-component school. By “downwind”, we mean the component that is receiving the energy or is more distant from the converter.
The free-floating school tends to believe that distortion of the metal caused by hoop stresses and compaction of the metal in the thread areas causes more problems than the stud will if allowed to float free (i.e., not torqued into the downwind component). There is tremendous merit in elimination of the stresses put on the tool metal by the inclined-plane action of threads, and I would certainly advocate this approach, except that I cannot imagine how leaving the stud to float free eliminates it. All things being equal, since the stud cannot logically be doing its job of keeping the mating surfaces in intimate contact without being in tension, the threads of the tool metal must be bearing a load as the result of this tension. The argument can be made that torquing a stud into a threaded hole so that it does not come loose creates additional unnecessary stress. The free-floating stud, however, is not without problems – chief among them being the unpredictability of the stud walking one direction or the other over time. Another issue is the possibility of sufficient relative motion developing between the component and the stud, which could cause the stud to become fused in the hole and render the tool unusable.
The torqued-in school is divided into two camps. One approach is to use a bottoming tap in the stud hole of the downwind component and a knurled-end set screw for the stud. The stud then is run down to the bottom of the hole and the knurls bite into tool metal at the bottom of the hole.
The other approach is to use a standard or semi-bottoming tap which does not bottom in the hole. The stud then runs into the area where the thread tails out and locks into the tool metal, as its first thread is constricted by the last thread of the hole. A cup-point set screw usually is used in this instance. Cup-point studs can be re-used as long as they are undamaged; whereas it is generally recommended that a new knurl-point stud is used any time the stud has been removed from the tool.
The knurl-point argument is that the extra hoop stress caused by running against the last thread with a cup-point stud to achieve locking is so great that tools fail at that point. The cup-point argument is that the knurl of the stud actually creates stress risers in the tool material, as well as placing extra tension on the angle created where the cylindrical part of the drilled hole meets the conical end of the hole.
While tools have failed in the stud area using both systems, the most common failure point using the cup-point approach is at the end of the hole, not the end of the thread (remember, this approach does not use a bottoming tap). The cup-point approach would seem to have the advantage.
The most important thing with any method is to follow the manufacturer’s recommendations about torque. “Old school” thinking usually errs on the side of over-torquing everything, but tight enough is tight enough, and too tight causes at least as many problems as too loose.
The single most important factor in keeping ultrasonic tooling joints working well for many cycles over many years is to ensure that the mating surfaces are as flat as they can possibly be. There is absolutely no way two conical surfaces or potato-chip-shaped surfaces will ever be able to effectively transmit high-power ultrasonic energy without significant heating and other problems. How flat is flat enough? The specifications vary somewhat by manufacturer and frequency, but rarely are surfaces allowed to be out of flat by more than about 0.03 mm (a little more than 0.001″).
Scored or grooved surfaces are sometimes usable depending on how much surface area is missing as a result of the scoring. The best approach is to send that tool off to somebody who knows what they are doing and have the grooved surface machined and the tool re-tuned as necessary.
From time to time, tooling interface surfaces may need to be lapped to remove surface deposits and mitigate pits and grooves. Most manufacturers’ recommendations on procedure are very similar, so we will not go into detail here.
Interface Treatment – Grease, Washers or Other Stuff
It seems no topic is more open to debate than how to treat the mating surfaces prior to assembly. One key point is that no liquid surface preparation is ever applied to the stud; but instead, is only applied to the flat mating surfaces. Various surface treatments appear here ranked according to the opinion of this author from generally worst to generally best. Not every rule applies in every case, however, so if you are having success with something and disagree with me, by all means, carry on!
Dry – no surface treatment at all. When high-power ultrasonic energy moves through the joint interface, it causes the metal surfaces to come into ever-more-intimate contact with each other. The more cycles are run and the greater the amplitude, the greater this effect will be. Also, since it is impossible to actually put every surface molecule of both mating surfaces into exact planes, there always will be some microscopic gaps that are potential sources of heat. Wetting the surfaces even a little bit will significantly improve energy transmission.
Skin grease – the oil from human skin. This sounds bizarre, but it has come and gone from favor several times in the last 25 years. It certainly is better than no surface treatment at all; but the fact that skin oil chemistry varies wildly depending on genetics, diet and time of year makes it hard to recommend. There are probably other organic oils (you could buy at the grocery store) that would produce more consistent results, but I don’t think I could recommend them either.
Light machine oil – approximately SAE 10W or lighter. In plants where silicone grease is banned, usually because of printing or painting operations, petroleum-based lube often is used. The case for light machine oil is that the film is automatically very thin. Frequent disassembly is required, however, because petroleum-based lubes generally will leave behind carbonaceous deposits on the tooling surfaces, which requires cleaning and lapping to remove.
Petroleum jelly – yes, you can find this in the “health and beauty” aisle. This has the silicone-free advantage of light machine oil but with much higher viscosity, so it tends to go on with a bit thicker (somewhat more forgiving) film and stays in the joint better. It leaves carbonaceous deposits like any petroleum-based treatment.
Metal or plastic washers – commonly made of Mylar but also copper and aluminum have been used. The significant advantage that washers have is that they are really easy to get right: put in one new washer prior to assembly. There is no technique to learn, no secret method, no skill – just put one in. This method is the most effective at preventing “fretting” (the build-up of oxides on the mating surfaces). They generally work very well and if you have trouble with one of the other methods, try this.
High-pressure silicone grease – usually Dow-Corning #4 or #111. The wetting action (improvement in sound transmission) of silicone grease cannot be matched by a washer. Silicone grease is inexpensive compared to washers, the grease generally stays where you put it and if the proper technique is used, nothing out-performs it. The proper technique is to simply get an unbroken film across the entire tool that is as thin as possible. Wiping as much off as possible using a bare finger usually leaves the correct amount on the surface.
Never put any liquid surface treatment on a stud! Never use any liquid surface treatment with a washer!
Once again, use of an appropriate tooling vise is recommended, because there will be several expensive cylindrical components on the workbench during this process and using spanner wrenches is often awkward. Its best to minimize the possibility of having one of those parts violently hit the floor (resulting in possible production downtime and a lot of expense). Also, tooling vises cause less stress damage to the “spanner holes” than spanner wrenches do.
The best assembly order is to clamp the converter in the vise and add the booster; then clamp the booster and add the horn. Obey published torque specifications. If you do not have torque spanners and cannot obtain them, then use regular spanners and do the math to figure out how hard you should press on the tool handle. It is probably less than you think.
Over-torqued joints may run fine when the tooling is young; but over time, the excessive compression of the metal around the stud will result in distortions of the interface surfaces, leading to excessive heating, which leads to more distortion. Some of this distortion can be corrected by machining the surface, but ultimately it is best to prevent it from ever occurring in the first place.
A properly assembled ultrasonic stack will idle at very low power with no noticeable increase in power draw as the operator continues to apply ultrasound for ten to fifteen seconds. The best indicator that something has been done incorrectly is a joint that gets warm after a fifteen-second test.
Twenty-five to thirty years ago, all of ultrasonic welding was seen as a black art, with so many variables unknown and uncontrollable that it seemed that the wingbeats of a butterfly in Tokyo could affect process outcome. With the coming of more precise and repeatable machinery, newer control methods and more science in tooling design, understanding of stack assembly is the last part of the process to crawl out of the cave and into the light of science. If sufficient care is taken, stack component life and process consistency will be improved, and that means less headaches and more profitable operations.
Tom Kirkland is a veteran of the ultrasonic industry, having begun his career as a customer over 25 years ago, and worked for a major plastics assembly machinery and tooling company for many years. He is a patentee in the field, has many published articles and technical papers and is a past president of the Ultrasonic Industry Association. Today he is a consultant in the area of thermoplastics assembly and is owner of www.tributek.biz, providing parts and supplies for ultrasonic welders.