Powder coating changing D2 toolsteels property? 2

[kekejg]

We powder coated some D2 Tool steel parts in two batches. The two batches were both heat treated to a hardness of 60-62 HRC. Both batches were double tempered at 400F. However the 2nd batch's fracture force dropped by almost 25% after painting, when the first batch did not drop significantly. We did various tests at the paint shop and believe we have eliminated the iron phosphate rinse as the culprit AND the tests would indicate that the baking WITHOUT the paint is not the cause which leaves me to question the “facts” seemingly demonstrated by our tests. If they are credible, then I would conclude that the cause is a chemical reaction with the paint and subsequent baking (we cannot test painted clips without baking because the paint would not cure).

The powder coating process on the D2 Tool steel consists of the following
- Bath in an iron phosphate solution 120F
- Fresh water rinse
- spray the polyester powder using an electrostatic gun
- Baking at 400F after powder coat (temp. can drift up to 425F) anywhere from 10 to 30 minutes

We had a couple of questions about the problem we've been having
What happens when the temperature if the oven drifts up to 425 F which is higher than clip's tempering temperature?
Could it be the powder chemically reacting with the D2 Tool steel at 400-425F causing the force drop?
What could be causing this inconsistency in fracture force drop?
Do you have any other ideas what might be causing this drop in fracture force?

 

[swall]

What test method do you use to determine "fracture force"?
What is the part configuration?

 

[TVP]

Have you had a proper metallurgical evaluation performed? D2 can have substantial variability in the primary carbide size, which has a significant effect on its fracture properties. There is definitely not a chemical reaction between the polyester paint and the D2 tool steel that is causing a fracture problem. Perhaps the phosphating operation is inducing some hydrogen into the steel and this is promoting a hydrogen embrittlement issue, but it is definitely not the paint. With regards to paint curing temperature, if it is higher than the previous tempering temperature, then there will be some softening of the tempered martensite microstructure, and possibly some dimensional change.

 

[metman]

With regards to paint curing temperature, if it is higher than the previous tempering temperature, then there will be some softening of the tempered martensite microstructure, and possibly some dimensional change.

.

I would expect this softening from "ovetempered" martensite to produce slightly higher impact results certainly not 25% lower.

TVP's suggestion of possible H2 embrittlement is much more probable.

 

[kekejg]

First our fracture tests are destructive ones in which we break the material across its width fracturing it through its thickness while we measure the applied force.

We tested both lots of clips after heat treat and before the powder coating process. Test results at that point were as expected. It was only after the powder coating process that the fracture force drop was evident and seemingly not on every batch. That is why we undertook testing after the iron phosphate rinse, after rinse and baking (with no powder coating) and the fully processed part. Based on this the conclusion that the powder coating/ baking part of the process is the factor affecting the material property. We had considered hydrogen embrittlement as a possible problem but tended to discount it given our second shop tempering trials on unpainted, heat treated material with no evident change in fracture force.

 

[unclesyd]

Me whole career has seen me working in some way or another with D@ tool Steel. We latterly have thousands of parts made from This material mainly for several reasons. growth while at our process temperature, its wear resistance. and ease of heat treatment. All our D-2 parts are heat treated the same, 1850°F air cooled, and tempered at the secondary hardening temperature. The tempering cycles are always two while we have gone to nine times for gear pump parts. We operate these parts at temperatures from 350°F to 685°F on 24/7/365 basis. The majority undergo periodic cleaning with temperature to 900°F.
I can't recall questioning any physical properties of the Heat rated product. We did do a little work with the heat treated directional properties of some shapes and there were a few minor difference in the traverse and longitudinal sections, but not enough to bother us. We also ran some reduced section Chappy Impacts on 2" thick end plates from older pumps that that cycled several hundred time.
You problem could be in the Heat Treat, but D-2 is very forgiving or whether the parts take from traverse or longitudinal section. Check how your parts are stacked for heat treating.

 

[JoeyV87]

It's possible that this is just a surface effect, and not a through-thickness issue. If you create an area on the surface which is effectively a crack (any surface without perfect coherency will have an effect), it will have a great tendency to propagate through the whole section under stress. The most useful information would be gained from a surface-breaking micro to look for evidence inter-granular attack.

 

[kekejg]

First, we use a flat ground D2 which we understand is cross ground. so grain direction is not known and we just assume that the grain is consistent and runs the length of our rectangular sheet.

I would be interested to know what going from 2 to 9 tempers does to the physical properties in your view. I would expect pretty much full transformation of any retained austenite which means more consistent behavior under stress. Beyond that?

As far as loading of the parts, we have experienced with heat treat shops all of whom ignore the requirements for the heat treat process. Currently we have just found out that our parts are stacked on edge with the shop's assumption that, that will have no detrimental affect on our part although they had agreed at the outset to hang the parts on moly rods. But because our parts are thin, i suspect this will cause more long lines of distortion rather than uneven heat soak.

Yes, propagation of micro cracking is a possibility but based on previous SEM, we do not believe this to be a case here. We come to the conclusion based on the fact this happens to a complete painted batch whereas micro cracking will be unlikely to be reflected in a consistent manner- that is the influence of micro cracking would cause a inconsistent results within a given batch.

 

[unclesyd]

The particular heat treat procedure for D-2 using the 9 tempers was worked out many, many, years ago by the Zenith Pump Company and DuPont in cooperation with Vanadium Alloys and Bethlehem Steel for manufacturing gears, gear plates and end housings. for some small metering pumps specifically for extrusion of Nylon Polymer. At the time there were many concerns about variable denier of fiber and gelliation of Nylon in the pumps.
The goal was zero growth during a cyclic operation at 600°F and occasional cleaning with exposure to 1000°F for several hours.
At the time when it came to retained Austenite it was the considered the same as the rabbit and hair problem. You convert 90% the first pass and 90% of the remainder and so on until the material will be stable at temperature. Remember that there was no LN2, cold boxes, and Swiz mountain tops close buy. If my recall is any good 5 tempers would get you by where the clearance's were at 0.0005", but to get to where there we no problems with smaller pumps with 0.0001" clearance ti took 7 or 8 tempers. Yes Zenith could grind this close and we could measure this close using Air Gauges and Micrometers from Curtis Wright.
When LN2 came along we ran numerous tests to verify any new heat treat procedure for these particular parts.

Remember at the time we were limited to 1000x optical microscopy as the SEM wasn't around, There was not any sophisticated electronic or ultrasonic measuring devices either. It was essentially heat and measure. What measurements were accomplished with what is now called a differential screw micrometer and some air gages.

I just checked and what few of the small pumps still in use some have serial number that they were manufactured in 1950s and are still in spec.

Another thing today the metallurgy is better and and the limits are tighter. As we always tempered at secondary hardening peak on the tempering curve we found that the Bethlehem D@ would consistently be 3 points that others.

 

[Robertmet]

Unclesyd , the old days , I remember them well ! Crucible recently made some CPM-D2 which has better properties by getting away from large carbides.I think this was made primarely for knife makers. I don't know if this will be continued with the sale of Crucible.
I agree that H2 should be investigated.

 

[etarney]

D2's toughness properties can vary widely depending on microstructure (degree of carbide banding or segregation) and orientation. If you purchased pieces of similar size there is no guarasnteee on teh orientation, and the impact properties could easily vary by more than 25% based on orientation. If you have broken samples, then you should be able to get a relatively inexpensive metzallographic exam performed to determine orientation, and check carbide segregation, between the two pieces broken before and after. If you did not cut them up yourself into pieces from a single larger piece, then I would not make any aaaumptions about the orientation.

 

[Wrenchbender]

Keke,
You previously referred to this part as a 'clip', mentioned rectangular 'sheet', and said the parts are 'thin'. A rolled sheet can have mechanical texture: highly directional properties. Are all of your clips oriented in the same direction wrt the sheet from which they are cut? That is one of the points made above by Etarney. If not, maybe the source of your problem is at hand. Perhaps the painted batch had more Ts than Ls. (?)