Pump Action Causing Micro-Bubbles in Polyurethane Finish

Pros puzzle over what causes tiny bubbles to appear in polyurethane finishes. December 1, 2005

We make a few exterior items. I'm looking for anyone who has used ML Campbell Euro X. I have the info from Campbell, but how does it really work? I've heard that the fumes made the fellow spraying it sick, physically. We currently use Sikkens and are pleased, but if there is something easier, cheaper, and better, I'd like to know.

Forum Responses
(Finishing Forum)
From contributor B:
I have been using this product for about 7 months on some projects that need polyurethane. The Euro-x is a very user-friendly product. I've used it on everything from Spanish cedar garage doors to a MDF kitchen.

The product likes to be retarded at 10%, no matter what the humidity level is. One needs to be extra careful when mixing the different coatings, as the catalyst for all the coatings is a different blend. They do make it user friendly by applying a color code that's visible on the cans, and the mixing ratios are also on the can. One doesn't need a spec sheet in order to mix any of the sealers or topcoats. The mixing ratios are all different, and I suppose that's the reason for labeling the can. Sealer is purple, clear satin is yellow, high gloss clear is blue, and pigmented satin is green. They all go over a clear sealer and can be tinted to any color with 844's.

The only problem is spraying the product with small component pumps. The down stroke is not equivalent to the up stroke, which results in air trapment due to the shearing action. I tried it on a 10 to 1 Kremlin and had a problem in the topcoat, even retarded at 10%. I used a Devilbiss kb2 pressure pot with #764 tip and the results were good. As is the case with all polyurethanes, they don't like to be pressurized. I will be doing some testing with a larger component pump on the next job, and I think that the more even up stroke and down stroke will eliminate the problems with the air trapment in the topcoat.

From contributor L:
You mentioned a shearing action that causes bubbles with some spray equipment. I have heard of this before, but do not really understand the how's and why's.

From contributor B:

I don't know the physics of why the shearing action creates porosity or air trapment of polyurethane, but I do know that it involves the upstroke and downstroke. The smaller pumps don't have an equal ratio of motion or pressure due to their design. On the upstroke, the ball lifts off of its seat and fluid fills the chamber above the ball. As the air motor cycles, the downstroke begins and the ball is seeded, pushing the fluid through the line instead of back out of the pickup wand. The characteristics of the polyurethane in this scenario may become turbulent prone with air bubbles being injected into the fluid. The fluid will be pushed through the lines out of the tip and not broken in the atomization process. The final result is micro bubbling buried deep under the 4 total mils of coating you just applied, where it will remain trapped after the polyurethane skins over. Retarding the product at 10% will help let some of the air escape, but there always seems to be some that gets buried too deep to escape. Some coating manufacturers urge you to not put coatings on a paint shaker, as it creates bubbles and foam, which can show up after you spray the product. I think this is along the same line as what's taking place inside your pump. I don't know if this is what happens, but this is my hypothesis.

From contributor G:
I'm quoting a contributor's post at another forum thread here:

"I questioned if you were applying too great of a film thickness, however 2 mils is not excessive. In most spray systems, there is a correlation between nozzle size, fluid pressure and the resulting flow rate. The viscosity of the material you are spraying may be high, so in order to get the required flow rate through a given nozzle size, you must increase the fluid pressure. Conversely, you can get the same flow rate using a larger nozzle size and a lower fluid pressure. Micro-bubble can be caused by sheer forces when a fluid exits an orifice at too high of a velocity. So a larger nozzle and lower pressure should reduce the bubble."

I know this can also be a problem when spraying solid colors. Somehow the pressure kicks the tinter out. So far as bubbles caused by shaking - I always thought that was one of those sprayer's myths. Does anybody have any real info on whether bubbles can survive being atomized?

From contributor B:
I don't think the viscosity has anything to do with this particular problem, because I only get this problem when I shoot polyurethane through a 10 to 1 pump. I don't usually have a consistency problem with any of the products I run through my pump because I use virgin material, catalyst, and retarder only. No thinners are used and my flow rates, material flow, and transfer efficiency are perfectly adjusted before the spraying begins. I check all coatings exiting the gun for temperature and viscosity through a zahn cup #2. The results have to be perfect to achieve consistent drying time and consistent sheen levels. The air cap is always the same and I use the same tip size for all the finishes that I spray a #09-154. This allows me to spray a box coat with consistent hand speed and the end result is 4 wet mils of finish. High solids or not, the products that I spray are orange peel free and because of the consistent drying time, dirt free. The only other problem I have is with polyester sealer, but most finishers have never even sprayed this, let alone tried to adjust a dual component external mix pump at about 1800 lbs. of nozzle pressure. It may be a myth about shaking clear coat finishes, but there must be some reason the manufacturer adds anti-foaming agents to their finishes. It can't all be about shipping or transit.

From contributor G:
Contributor B, you are outside my range of experience there. What does your supplier have to say? Please keep us posted, as the esoteric problems are the most interesting ones.

P.S. The defoamer is to prevent foam. I don't know that foam carries through the gun.

From contributor L:
Thanks for your insight. It does make sense that turbulence can be created within the fluid section of a pump if there is an air pocket, and equally when shaking some finishes. These bubbles probably would be across the entire substrate, not necessarily isolated to the deeper grain areas of oak, walnut, etc., as would be associated with air entrapment. Would that be a safe assumption?

From contributor B:
Contributor G, most suppliers are salesman, not engineers, and most of what a salesman states is hearsay. I have asked a few people about this topic and they proved how to eliminate the problems with shooting polyurethane through a 10 to 1 pump. We came up with the idea of using a larger pump to eliminate the problem.

The reason chemical manufacturers use anti-foaming agents is shipping and transit of materials. Most coatings are delivered via common carrier trucks. The product gets shaken up in transit and foam accumulates within the space at the top of the can. If it weren't for the anti-foaming agents, the foam would remain constant instead of settling back into its original state. I don't think anyone has to worry about foam going through their guns.

From contributor B:
I experienced small patches of holes and bubbles on two different bar tops . The mahogany top had 15 mils of polyester sealer and the cherry top had 8 mils. I know that the holes were not from holes in the sealer being open and the topcoat not being thin enough to fill the holes due to the bubbles. Due to gravity, the bubbles rose to the surface popped and weren't able to refill the hole that was left, even with retarding. Reshooting wet on wet to fill them would have led to a large number of mils, and clarity problems, perhaps even adhesion or cracking problems at a later time. I have seen the bubbles on substrates of deep grain products such as oak and black walnut, of which you speak. This is really not the same issue as the ones that you describe. One is bubbles from substrate substance and the other is made from machine. If you have the problem you're talking about, just increase or add retarder.

From contributor R:
Let's think about a couple of things here.

1. Shaken and not stirred. Most spray equipment atomizes finishes by injecting vast quantities of air into them, so what do some bubbles in the can matter? Brushed on varnish is usually recommended to be stirred and not shaken because its high viscosity would not let the bubbles rise to the surface and pop. I may be wrong, but anti-foaming agents are generally used only in waterbourne finishes. There is a difference between foam and bubbles.

2. Micro bubbles. If the pick-up tube of the pump is submerged in liquid, where does the air come from to make the micro bubble? If you picked up air from the tube, you would lose the prime on the pump.

Most AAA guns optimally atomize material by producing 400 to 600 psi of fluid pressure at the tip. The sheering action takes place as the liquid exits the nozzle of the gun at too high of a fluid pressure. In effect, it is over-atomized. The droplets of finish it is producing are too small. As those tiny droplets land on the wood's surface, they have tiny air spaces around them. As the droplets flow together, some of the air in these tiny spaces gets trapped, producing teeny tiny bubbles. These tiny bubbles are not very buoyant compared to the viscous liquid they are suspended in, so they can not float to the surface fast enough to break, and thus they get trapped in the layer of dried film. A larger droplet will have larger spaces around it, and thus produce bigger bubbles, which are able to float to the surface faster, break and re-heal.

A 10:1 pump mechanically works the exact same way as a 30:1 pump. The two differences are that the 30:1 produces 3 times the amount of fluid pressure at the tip for every pound of input air compared to the 10:1 pump. Also, the 30:1 can produce a higher sustained fluid delivery rate, in ounces per minute, as the 10:1. Why is this important? It's all about the flow. You want a certain flow rate in order to lay down a nice wet coat in a reasonable amount of time. As the viscosity goes up, so does the required flow rate. 2k poly has a higher percentage solids by volume ratio than, say, a pre-cat, so it will also have a higher viscosity.

You can increase the flow rate two ways... Increase the fluid pressure or increase the nozzle size. If you increase the fluid pressure too much, you run the risk of micro-bubble (as per above), so the alternative is to use a larger nozzle with a lower fluid pressure behind it. You could use a 10:1 pump, but a 10:1 pump with a lower fluid delivery rate (ounces per minute) will be working like a dog to keep up with the demand of the nozzle, whereas the 30:1 will barely be breaking a sweat. Finally, you could reduce the viscosity of the material and go back to a smaller nozzle.

When you went to the larger size pump, did you check out the size (flow rate) of the nozzle in it compared to the 10:1 pump?

From contributor B:

You indicated that air bubbles are injected into the finish as it passes through the lines to be atomized. How could that be, when your air source is totally external? These are air assisted airless, and that means that you turn your pump up to where you get a nice stroke and an even flow coming out the nozzle. You then turn the air up while spraying to make the horns disappear. This is how you adjust the pump to spray. As indicated earlier, we are an airborne solvent free shop. This means is virgin material, catalyst, and retarder only. The only thing we use solvent for is to clean the gun after we're finished for the day. My viscosity is always the same as what's recommended by the manufacturer and measured with a zahn cup #2 and a temperature gun.

You indicated that you couldn't understand how air bubbles could form out of a nozzle. The pick up tube running out of fluid is one of the options, but is not what's taking place here. The shearing action or pockets of air are created at the top of the stroke or the bottom of the stroke, when the movement is uneven at this point . The larger pumps are designed to have equal motion on both strokes and that's the reason they work. As I indicated earlier, my viscosity is not the problem and it really has nothing to do with my air pressure and fluid tip sizes. If I use more air pressure, more pump pressure, and a larger tip, then I can see my transfer efficiency going right out the window. The object is to paint the product, not the ceiling, walls, and floor around the product. I use the same tip for all my finishes on both pumps, 10 to 1 as well as 30 to 1. How does one reach the recommended viscosity on all finishes without using thinner? Do you use catalyzed products? What do you do with your leftover catalyzed product at the end of the day? Do you throw it out ? Do you seal all exposed wood on what you spray, seen or unseen?

From contributor R:
I don't believe that the air bubbles form inside the hose, but rather at the point where the material exits the nozzle and meets the outside air. I am looking at the schematic for Sata pumps and mechanically, the 14:1 is the same as the 32:1, just a smaller air motor and piston. I can not speak for other manufacturers designs. My point is that you use a larger tip size and less air pressure. That is what will increase your flow without the micro-bubble.

Viscosity is a function of percent solids by volume and temperature. So if you need to lower your viscosity, you either add a reducer or raise the material temperature. Most manufacturers state the viscosity of their material in the can at a certain temperature - Campbell uses 77 degrees. The viscosity required is by a particular piece of equipment, with a specific nozzle/needle/air cap, usually different than what is stated on the can. You need to adjust the viscosity to fit the piece of equipment you are spraying it with. Most people do this with solvents, since heat systems can be complex and will shorten the pot life of catalyzed products.

I shot 2k polys almost exclusively for over 12 years. Many brands through most types of equipment. I let the leftovers harden and then had it hauled away with our liquid waste. Our standard procedure was to seal all surfaces.

From contributor D:
Back to the original question, let's take MLC at their word that their exterior-use 2k poly is worry-free for at least five years. Great. What happens after that? The customer is calling up and needing a new finish. What do you do then?

Many customers' needs are better served with a lower-rated coating that can be maintained. Maintain the coating before it degrades, and you are good to go until the next maintenance call. It's good all around - happy customers who pay for maintenance contracts, and finished items which can be revived without having to strip off the five year old "sherman tank" coating.

As for interior use, I bet the MLC 2k PU is swell. But did I hear a quote of up to $80 a gallon including the sealer? Ilva weighs in at about $26 a gallon. Um...

From the original questioner:
I just heard from Campbell that absolutely, when it needs refinished, you've got to take it back to bare wood. Yikes.

The comments below were added after this Forum discussion was archived as a Knowledge Base article (add your comment).

Comment from contributor P:
I think the problem is related to the isocyinate hardener and high humidity. As water laden air travels over a wet film during film formation, water reacts to the isocyinates in the hardener causing a CO2 explosion which results in tiny blisters or bubbles in the film. I studied in Germany with a German automotive paint company and all of your modern automotive topcoats are basically the same technology.