SUGGESTED WELD PROCEDURE 1

[FldEngr]

Our shop is welding a 1/2"x 6" wide TP-100 steel plate (spokes) to a circular (tire) ring, 1030 CS, Q&T to 90ksi, 5" thick x 11" wide x 8.5' diameter. There are two rings, one at each end of a circular drum for supports. The drum rotates at approx. 8rpm's.
The 1/2" plate is welded to the ring via two oval holes approx. 3" long x 2" wide with 2 to 3 weld passes - the weldor does not fill the oval with weld completely. We are using MIG wire, 90ksi with shielding gases. Currently, the shop is preheating the 5" thick ring to 150 deg. F. with rose-buds and maintaining that until welding is complete. There is no post-weld heat treatment except for cooling to ambient temp.
We are seeing some in-service failures of the TP-100 plate sometimes at the edge of the circular welds and sometimes approx. a couple of inches from the two circular welds.

Questions:
Should there be a different preheat procedure?
Should there be a postweld heat treatment?
Are the any other suggestions for making this particular weld?

Thanks very much for your input.

 

[metengr]

FldEngr;
What is TP-100 steel plate? Do you know the chemical composition and heat treatment requirements for this material?

Regarding the cracks near the weld attachment points, it could be from several factors - inadequate preheat or poor weld design based on the attachment of the TP-100 ring to the 1030 CS substrate. Is the purpose of this outer ring of TP-100 to serve as a wear resistant surface or what is the function of this outer ring?

 

[stanweld]

You have not provided much detail regarding the failure mode or the general lifetime to failure. If the failure is short term fatigue, did cracking initiate at weld flaws, or at geometry changes at the surface(s), in weld metal or HAZ? Apparently delayed (hydrogen)cracking is not a problem.

 

[TVP]

TP100 is a designation used by Steel Warehouse for a HSLA steel with C = 0.10 max, P = 0.025 max, and S = 0.006 max. The yield strength is minimum 100 ksi, UTS = 110 ksi min, elongation = 15% min in 2 inch gauge length. They used to source this from European mills like voestalpine, since there were no US mills producing this type of HSLA steel. Not sure if that is still the case.

 

[Tmoose]

How many hours/cycles does it take to break something?
At 8 rpm, 24 hour service = 11,000 cycle per day
How many spokes? more than 2 I hope
Any forensics done on the cracked parts? should be
Any pix available?
How is the drum affixed to the (tire) ring?
How warm does the drum get? How often does it cool off?

I don't have any mental image of the part or the welded joint or the forces, but I suspect this may be less of a welding problem, and more of a part and weld geometry problem. geometry > stress concentration .
constrained parts

Straight spokes in cast wheels and pulleys were long discouraged. The inevitable shrinkage stresses are just too great, and easily minimized by offset in 2D or 3D.

http://www.tooltimer.com/images/barnes3-2l.gif

http://www.leesidetools.co.uk/boxford lathe211204.JPG

Although I have no idea what the orientation of the weld is, the description as an oval hole sounds like an admirable effort to minimize weld running "across" the alternating tensile forces. Omar Blodgett has written many articles about the dangers of transverse welds subjected to tension. The latest issue of WELDING magazine (not onlne yet) has one article. many of Omar's articles are published here.

http://weldingmag.com/blodgett/

think the 3 toward the bottom of the first page are about ductility are especially good

 

[RPstress]

Many thanks to Tmoose for that link to Olmer Blodgett's pages. I'm amazed he's still alive, yet alone doing a sort of engineering blog!

I first came across diagonal bracing (

http://weldingmag.com/blodgett/no_twisters_swingers_allowed_0501/

) in his book in the late eighties, and very enlightening and useful it was too.

 

[autiger74]

I apologize for the poor equipment description. This is an asphalt mixer drum; i.e. an 8' dia. drum x approx. 30' long and supported by these "tire/spoke" units - one on each end. The spokes are welded to the inside dia. of the tire and the outside dia. of the drum shell. The spoke has a "break (bend)" approx. 8" from the end and this end is attached to the I.D. of the tire via two oval welds.
The drum side of the spoke is also attached to the drum shell with two oval welds. Typically, there are approx. 20 to 30 spokes supporting the drum inside the tires. The tires rotate on trunnions and driven by a heavy duty chain drive.
Sometimes a few spokes may crack within the first year of service and others may run several years before ever showing signs of cracking. The drum rotates approx. 8 to 10 hours daily all year.
I have a "gut" feeling that there may be a factor with the "break" in the spoke at the tire connection which is only inches away from the weld to the tire and possibly the shop's preheat/postheat treatments.

Thanks again,

 

[stanweld]

I tend to think that design geometry is the culprit rather than the welding procedure' preheat or PWHT. Fractures in the spoke inches from the weld would not appear to be due to welding residual stress. Have you considered peening.