Clevis pin fails double shear....Is rework possible?

[MCFcorp]

I have a lot of clevis pins that failed double shear test. I am trying to determine if they can be reworked to an acceptable condition. The part is MS20392-12C117. It is a 1" diameter by 3.6" effective length pin, cad plated to AMS-QQ-P-416 Type II, Class 2.

Here are the details as the lot sits today;
1. Material 4130
2. Heat treated to 26.5-32 HRC iaw AMS-H-6875A
3. Double shear to be 116,260 lbs min.
4. Independent lab machined specimens to 7/8 dia to fit their fixture. The double shear requirement for 7/8 is 88,840 lbs.
5. 13 samples ranged from 85,544 lbs - 89,118 lbs. only 1 sample was conforming
6. Heat treating facility certified parts to 30-32 HRC
7. My hardness verification shows 24 - 28 HRC w/ 50% of the samples non-conforming.

The question is...
Can these parts be reprocessed to achieve the required properties?

I think I should be about to anneal or normalize the parts as defined in AMS-H-6875 then reprocess through austenitizing, quenchin, and tempering.

What is the correct process...Anneal or Normalize?

Thanks in advance for any insight

 

[IRstuff]

Did the machining alter the properties? What makes you think the same thing won't happen again?

Perhaps you should find someone who can do the test without messing with the material.

TTFN

FAQ731-376

 

[MCFcorp]

IRstuff,
Doubtful, I had thought about that but they alter test specimens all the time. It is a fairly common practice in the fastener industry for mechanical testing.

The bottom line is my hardness results are deviating by as much as 30% from those listed on the heat treaters certification.

I am confident it is a process issue at the heat treaters.

 

[IRstuff]

Have you taken your stuff back to the supplier and witnessed some retesting? Seems to me that you're arguing that they've essentially lied on their certs, and you should demand that they do a retest with a witness present to demonstrate compliance or not.

If their testing also shows non-compliance, then it's up to them to make it right, whether by reworking or replacement. Either way, it shouldn't be you doing the reworking.

TTFN

FAQ731-376

 

[MCFcorp]

I am already scheduled to witness them retesting the parts tomorrow. We are pretty much on the same page. It is possible that the samples they pulled in their test were OK but that other parts of the lot were bad.

I just need to have a recovery plan and understand my options when I go there tomorrow.

 

[dbooker630]

I would also have a representative part sectioned, polished and etched for microstructural analysis. Is the heat treater capable of performing this? There can be several reasons for the discrepancy that are not due to misinformation.

Sampling-is their cert based only on a first or last piece?
Load density-your part did not get to austenitizing temp.
Quench severity-part shows slack quenched (bainitic) structure, quench temp fluctuation?.
Decarburization-was there a furnace issue?.

It helps to know the microstructure so the correct rework recipe can be implemented. Otherwise you're just guessing.

 

[MCFcorp]

dbooker630,
That is a excellent suggestion. I just returned from the heat treater. The hardness results from their facility were chaotic to say the least. The oven charts revealed that the tempering was only held for 45 minutes. The temp was 1075 F. The parts took roughly 1/2 hour to come up to temp during tempering.

There was a lot of sanding off of cad plating, dusting off the indentor, wiping off the anvil, wiping off the part, etc. to get acceptable readings. Even then the results were at the low end of the acceptable range. While non of the results reached those specified on their certification.

They cross sectioned a part and checked mid radius and core. Those readings were in the low 20's (21, 22) for both the mid and core. I am reasonably condifent that those are not adequate results. I will have samples sent for microstructure evaluation as you suggest.

 

[dbooker630]

MCFcorp,

Based on your description of the heat treater's hardness testing I would employ a third party to perform a micro for you.

Water-quenched 4130 1" diameter, typically has a hardness of 26-27 HRC after tempering at 1100F; 31HRC at 1000F. With the short tempering cycle you report the hardness should have been in the low 30's HRC - unless they are quenching in oil.

4130 is typically a water-quenched grade of steel. The micro will tell you if the quench is severe enough.

 

[MCFcorp]

they oil quenched at 144 F

 

[dbooker630]

For the moment I have my doubts about a sufficient quench but I'll wait for your micro results. If you had said 144F polymer I would have less doubt.

 

[redpicker]

With oil quenched 4130, I'm not surprised with these results. You might get marginally acceptable surface hardnesses, but the core hardnesses can be low. It's not impossible to get good results in oil, but it would require careful fixturing of the parts and a well agitated quench.

This also explains why your test failed. Machining the 1" diameter to 0.875" removed the hardest part of the pin.

These could be slavaged with a quench and temper operation (normalizing would not help much at this point). Water quench would be preferred, but if they are using an integral quench furnace, it may not be possible. If they have to use oil, have them fixture the pins in a vertical position with at least 1/2 radius between parts to insure adequate oil circulation during quench. The tempering time needs to be sufficient so that all parts see the same soak cycle.

rp

 

[TVP]

Going back to your original question, yes, re-working is possible to obtain higher strength. A martensitic microstructure is optimal for austenitizing, so just re-heat the parts to 815-870 C (hardness will be maximized by using the higher end of the range), then quench. dbooker630 and redpicker provided some good information about the various quenchants-- 4130 does not have great hardenability, so quenching into a water-polymer mixture will result in higher hardness. If this is not a possibility due to risk of distortion/cracking or unavailability, then use the "fastest" oil possible. If your heat treater does not know what fast or slow oil is, then you need a new heat treater. Tempering info was also provided previously.

 

[unclesyd]

While waiting on the microstructure results I would try to reprocess a small batch of pins.
Reprocessing will present a few problems since your parts are Cad plated. This plate will have to be stripped, preferably in a non acid bath.

I would normalized the parts prior to the Austenizing step.

This is almost a necessity if there is a ring or tab on the end of the pin.

I would austenize at 1575F and warm water quench and then temper at 1000F. Myself I would double temper.

After the test batch of pins are HT I would give them a wet PT examination.

 

[MCFcorp]

The lab still has the 13 samples from the double shear. I ordered microstructure evaluation on 3 of those samples. 2 samples are those that had the lowest load at failure during the double shear, the 3 one is the only sample that passed double shear. I should have results in 2 days.

In the meantime I may have half a dozen samples reprocessed.

What would be the difference between reprocessing as TVP describes vs unclesyd's recommendation to normalize? The clevis pin does have a head.

 

[CoryPad]

The differences would be cost and homogeneity, with unclesyd's recommendations possibly providing more homogeneity at higher cost. If your supplier is not doing the current process correctly, then they won't do this proposed one well either.

Regards,

Cory

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[MCFcorp]

higher cost isn't much of a concern. I have 15K in sellable product stuck in MRB. Making these good parts is the only option.

 

[mfgenggear]

MCForp

Being in this same situation .

This is a long thread all ready.


before doing anything I recommend this
I would recommend a driffrent lab to reverify met lab
properties. ( holding 2 points on the hardness scale to my experience is hard to hold)It may have been over tempered, or tempered to high.
In addition 1/8 material removal can make a big diffrence,if I under stood you corectly


#1) what the qty of parts ?

#2) are are parts centerless ground or turned?
what is the tolerance of the OD?

#3) if you reheat treat these pins there is no doubt in my
that it can be reworked to achieve the correct properties.

#5) however the question #2 applies, parts will distort
violating Straightness & or Tolerance requirement.

#6) supplier who did the testing machined the 1.000" pins
down to 7/8 dia? that is bogus.
according to my Timken Handbook 4130 hardness will rapidly drop.

 

[MCFcorp]

mfgenggear,
1. There are approximately 630 pcs in the lot
2. The parts are turned. The tolerance is +0.000/-.002
6. Good point. I will request that they retest using a proper fixture. If they cannot I will call and find another A2LA lab to do the testing.

What section in the Timken Handbook are you refering too?

 

[TVP]

MCFcorp,

The Timken Handbook can be downloaded as a .pdf file using the following link:

http://www.timken.com/en-us/Knowled...Practical_Data_Metallurgists(4023)-Sept06.pdf

I believe the specific information to which mfgenggear was referring is the 4130 end-quench hardenability data shown on the following page:

http://www.timken.com/en-us/Knowledge/engineers/handbook/Pages/pdm053.aspx

 

[MCFcorp]

TVP,
I appreciate the assistance as well as that help from everyone else who has spoken up in this thread.

I will update with results from microstructure and full size double shear test in a few days.