Two relatively easy low-tech approaches:

1. Weigh the relevant quantities, such as before and after adding the coating.

2. Measure total thckness at any number of desired sampling points, and subtract the thickness of the base material.

The coating thickness range you mention should be reasonably detectable.

Tornado,

Thanks - weighing is not particularly appropriate for a coating as you will only calculate an "average", not locate highs and lows. Measuring the overall thickness is fine for "simple shapes", but when it comes to a complex shape, (think multiple curves) the total thickness may not be easily determined by either UT or physical measurement.

Indeed, this does complicate the problem. If there magnetic or ultrasonic sensors that can detect the boundary between the coating and the substrate (I am not familiar with the possibilities), one would still need to take many samples. One way out of this might be a means that controls the spraying process so that it covers equal areas in equal times. If the spraying device overshoots the boundary of the object, there might be a way to deduct that weight from the weight that went where intended.

Thanks - the process is manual, and deliberate overspray is "expensive" (but may be a valid option). I have asked the manufacturer's of several "leading coating thickness gauges", whom have a inventory of both magnetic and ultrasonic sensors based gauges for measuring coating thickness for a great variety of coatings over different substrates. Unfortunately for me, I fall nicely into the gaps of what is not covered - i.e a thin semi magnetic coating over a magnetic substrate - magnetic gauge will not get the difference, and too thin / not differentiated enough coating for UT, and eddy current won't work.

That is why I thought I would cast my net wider to this forum. I have asked the same question on a more tightly focused coatings / surface engineering forum, and also struck out..

I came up with this overspray idea on the basis of the spray-painting technique of depressing the trigger just before encountering the edge of the object, and then movng just past the other edge before releasing the trigger. Rather than waving the can or gun around over the object, piling great thickness on some areas and all but missing some others. This is a bit of a guess, but I suspect you will need to traverse the object in parallel paths, with a bit of overshoot. For instance, close zigzagging would overcoat the areas where part of one spray path overlapped the previous. On the other hand, a circular disk could be sprayed in a constant-velocity Archimedean spiral, with very little overspray.

If I get the chance, I will play around with the idea of a serpentine spray pattern of traversing the object, making a U-turn at some "best" point, traversing again, etc. There would be some nonuniformity at the hairpin turns, but it still might meet the thickness tolerances, and still minimize wastage.

Cute little problem, this.

Do you know the answer to the question, or not? Jeez...

be nice!!

Is this a quality check? Or something that must be done to every part? Can you make a mock-up of the substrate out of a nonmagnetic material and spray it in a similar manner? This would at least give you an best-guess at your general thickness. But if you are looking at each individually sprayed part I am stumped for now.

I have used magnetic thickness sensors but they seem to have been thoroughly debunked as useful already. Hmm, will ponder this.

-T

Hi Kiwi, Tornado got it right, I worked for company that would spray monel onto ESP's and that how they did it, simple and very accurate.

Hi,

I would try ultrasonics.

If you are using inclined coupling (near 1 to 5° to plane of material) and elevated frequency (MHz) then there should be a pretty clear signal. Resolution may be an issue.

RHABE

Assuming you have access to both sides of the part, and that the substrate thickness is fairly uniform, you might try a dial indicator, either mechanical or optical. If you have access to only one side, I suggest an ultrasonic thickness gauge. These gauges work by measuring the transit time of an acoustic signal launched from the surface of your item. You will be measuring the transit time of the signal reflected from the substrate/monel interface. The amplitude of that signal will depend on the acoustic impedance mismatch between the two metals at the interface. If the match happens to be very good, there will be no reflection until the signal reflects from the far side of the substrate, giving you the overall thickness at that location. In ether case, you can obtain the information you seek. You must use a test/transducer frequency suitable for the thickness you are measuring. The applications people will help you with that.

I have done this using UT. You need to set up operating procedures using calibration blocks sprayed with the same substrate material and coating with the known spray parameters. Using these calibration blocks determine the various thickness and set them into the meter using a template (go or no go within a prescribed range) then apply this to your coated parts. It takes a little work, but once you have established the procedures, its flawless. If you are worried about the overspray, then by a robot...the payback will be less then a year. You can also sell the overspray from your "dry cartridge dust collector" to offset the cost of your feedstock.

P.S. The UT guys don't have a clue how to do this and be sure "not" to use couplent when using this approach. It will contaminate the coating and believe it or not, produce less accurate readings.

Hi,

Monel K500 is nonmagnetic!

316SS is minimum 10% lower in elastic modulus (E) compared to carbon steel at nearly the same density.

Most Monels are higher in density (ρ) (compared to SS316) and lower in elasticity.

As elastic impedance and thus detectability by ultrasound is square-root(E/ρ) both should be detectable, but Monel will be easier.

If the contamination from the coupling fluid is a problem (see above post by NICROMO...) and air is not sufficient, then try water or glycerin or water-sugar solution, all these can be cleaned by pure water. Water-sugar mixture can be made in any viscosity if mixed hot, used as candy.

If signal from the back is too high then my above-suggested method of inclined sensor-heads will work, or a back-surface of high damping and high density may work.

RHABE

Remember that you are measuring a thermal spray coating. It does not have a crystalline structure and does not behave like wrought materials.

You are right,

density will be likely lower than bulk and elastic modulus higher as partially oxidised and streessed, this will counteract my prognosis that was based on bulk data.

RHABE

You can use the UT B scan method or magnetic resonance lift off ( I think it is called) to determine the thickness of base metal and cladding. I am not sure these methods will work with Monel but I think they will. My contract NDT company uses these methods of to take thickness readings on 316, inconel, 312 stainless steel weld cladding on carbon steel material base metal vessel.