17-4 PH 2

[Ewan FI]

We manufacture a Measuring Axle from 17-4 PH. We buy in the material in Annealed state and machine this in house to the form we require which approximates to a 350mm long x 35mm dia "dog bone" shape with the center being machined to a 15mm square cross section. We leave a couple of diameters slightly large for post heat treatment machining.

We then sub contract the heat treatment to condition H900 and we are finding the Axles are bent when they are returned and hopefully someone on here can help me find out why.

Our process-
1. We measure (on a center) and record dimensions before we ship the Axles so we know they are good when the are dispatched
2. The sub contractor then anneals the Axles (not sure if this is necessary and I suspect that this is where the problem could be). They say this is necessary to achieve the hardness?
3. The sub contractor heat treats to 900 deg F with the axles vertical for 1 hr and then allows to cool in still air
4. We receive the Axles back and check the dimensions - the last batch had a 1mm runout!

From my reading up on 17-4 HP I can find that 900 deg F would not cause any deformation of the part (other than the acceptable growth which we allow for). Would the annealing temperature cause deformation? Would the annealing relieve stresses built up during machining and this is what I am seeing?

Any advice appreciated.
Ewan

 

[Dhurjati Sen]

Why would you carry out failure analysis on behalf of the sub contractor? Penalize and blacklist him for good.

DHURJATI SEN

 

[Ewan FI]

I'm sure they would argue that the failure is not their problem. We have asked them to heat treat to harden and that is exactly what they have done.

 

[Dhurjati Sen]

It is easy to catch them. Send a small batch and they have to carry out all the HT in your presence. Once the axles come out of annealing get hold of them and check the dimension. Cool !

DHURJATI SEN

 

[Ewan FI]

That is certainly part of our plan but I was looking for some advice from someone familiar with 17-4 HP. Someone's thoughts on whether the annealing is necessary.

Or plan is to send them a batch and have them anneal half and return them for measurement. Assuming they are OK we will return them to have the entire batch hardened and then test hardness to see if there is any difference. This will obviously take quite a bit of time, hence the reason for asking advice.

 

I think you already know the answer. You received raw material in the annealed condition, and there was no reason for the heat treater to repeat that. Your machining allowances work for the 900°F hardening stage but the part cannot be expected to survive annealing dimensionally intact.

"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[EdStainless]

What size is the starting bar? It is likely that the core of that bar is not annealed well enough to actually reach the H900 (very low toughness and easy SCC) condition. And this is likely why he is suggesting re-anneal.
They should be checked and re-straightened after the anneal. Are they doing this vertical also? They should. They will need to cool fairly quickly, but nothing like a quench. Non-uniform heating or cooling here could be an issue.
Non-uniform stresses in your original bars could also be part of this.
I would expect you to have to re-straighten these when done. That doesn't surprise me at all.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

EdS,
Thanks for educating me on the intricacies of stainless steels (not for the first time ).
Perhaps then an intermediate stage of re-annealing the bar in a very rough machined condition would make sense? This would mean an extra trip to the heat treater of course.
I agree holding the part vertically is best, but I have reservations about mechanically straightening these parts.

"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[Ewan FI]

I have just done a bit more searching on 17-4 PH on the forums and one post states that the material is annealed as a large coil, straightened and then chopped into lengths. This could add some work stress to the raw material.

Ironic metallurgist -
Thanks. I tend to agree (with my limited knowledge) that the annealing is the problem and is not required. Our proposed trial should prove/disprove this.

EdStainless
Thanks. Our raw material is 1 1/4".
Our drawing does not ask for the parts to be annealed prior to hardening and I only found that our supplier does this when I phoned them. The drawing does state that the parts are to be vertical during heat treatment so I would assume that they also do this for annealing (I will ask them this specific question). They say they use a wire frame to support the top of the Axle in the oven.
The drawing also states that the parts should be cooled in still air so cooling should be uniform.

The bent axles I have received can be straightened but I think annealing these prior to straightening will be required. Then re-hardening. We apply strain gauges to these axles so any inbuilt stresses will give false readings.

 

[EdStainless]

I would guess that 1.25" is annealed well enough to allow direct aging. Though mill are notorious for under annealing material.
How far above your minimum properties are you getting? If you could bump your aging temp up a little (925F) it will lower the strength and increase toughness a lot. Using a double age is also possible, age at 925F, re-straighten, then age at 900F for some stress relief.
By the way these aging temps should be controlled +/-10F through the zone and process.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[Ewan FI]

I have just had an interesting phone conversation with a supplier who has supplied us complete Axles in the past. Strangely enough they have used the same heat treatment shop and have had problems with them overheating parts.

It was suggested that we machine the axles from 17-4 PH H900 and not the annealed bar - we could skip the heat treatment altogether. Can anyone see any problem with this?

 

[EdStainless]

You need to do a test. The properties of the core may be different than the full bar either due to variation in annealing or actual microstructure.
I would suggest machining some standard tensile bars (12.5mm) from the center of your 35mm bar. Then have a few of these annealed, and then age all of them together. See if you get a difference (look closely at elong and RA).
Machining this stuff in the H900 is difficult but can be done with enough HP.
Is fatigue an issue with these? Do you peen them after machining?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[Ewan FI]

We have a plan that should hopefully give us a better understanding of where the bending occurs in our manufacturing process. I suspect we will find that it's the annealing as pointed out to me but time will tell!
This will involve us machining another batch of Axles and varying the process so will probably take a month.

I promise I will come back and post our findings and results. I search forums often to find solutions to problems and find it infuriating when people do not post up conclusions!!

Thanks to everyone for their input and for sharing your knowledge.

Ewan