Dale,

I thought that U were busy at work....

My only 2 bobs worth would be to use a backing strip behind the end dics to provide a better root for the weldment.

I think that this used to be done before on pulley designs but stopped due to cost and thinking that no required if welding can be done without it.

Cheers

James ■

Dear Mr. Coward,

Kindly let me know the thickness of the disc and the shell plate.Cocave welding can be achieved by using contact type of electrodes.I can give you complete weld procedure if I know the thickness of the plates,Material specificationand the joint configuration.

Regards ■

Dear Mr. Coward,

We can solve your problems if we know the Th. of materials,material specification and the joint configuration.We have already solved this type of problems.

Once we get these information we can give you complete weld procedures to achieve good results.

If possible if can send the sketch of your present joint preparation,it will help us to understand the problems easily.

Concave welding can be achieved by using contact type of electrodes.

Regards ■

Dear Dale,

Conveyor Dynamics, Inc. has extensive knowledge and experience with the type of failure you have described. We have predicted failures, with this detail, of operating systems, that subsequently failed. I think there is a photo illustration of the failure type on our website: www.conveyor-dynamics.com

Further to this we have developed a design criteria and speciification on the design configuration and triaxial fatigue stress for the pulley assembly including the shell to end disk connection details and stress analysis procedures. The degree of sophistication is dependent on the loads and metal thicknesses.

The shell and end disk should be thickness compatible within 2/3 of each other for proper weldments or the end disk tip is machined to give good heat transfer balance.

Most manufacturers cannot properly stress analyze your configuration with FEA or other methods. Your pulley cannot easily be described with the actual geomtric confiuration at the weld interfaces with the end disk and shell. You have made a relevant comment on the same.

This type of failure comes from the shell's hoop or tangential and bending stress having to morph to a localized high shear stress as it turns the corner from the shell into the end disk. At this point, the inside shell to end disk configuration is critical in the form of local stress risers or discontinuities. The inner weld or lack of weld does create unacceptable irregularities when the finish is not refined.

The acceptance of machined or cast end disks with T-type shell connection is one answer to the problem. There are a number of ways to control the stress riser (Schmotlzi Discer., Hannover Univ.). I will not dwell on the many ways that have been tried. Always, the economics plays a part.

Most manufacturers have tried to invent ways to minimize the discontinuity without reliable success such as: backing bars, ceramic dams. Often their answer is make it bigger. Proper inner weld and surface finish can be made to work. The butt weld external to the end disk is the most reliable. This assumes the end disk is > 50 mm.

Single sided weldment is the worst configuration. The internal weld end is highly irregular and has the highest stress riser.

We require grinding the path of a concave fillet weld on the disk inner surface, usually by submerged arc to a 125 RMS (english units). The final grind must have a concave shape to allow normal stress flow. The weld ends to disk and shell circumference cannot have abrupt discontinuities such as undercut, path irregularity, start-stop action pooling and undercut.

Good Luck,

Lawrence Nordell

Conveyor Dynamics, Inc. ■

Dale:

Here is the shell to end disk weldment failure photo referenced above:

http://www.conveyor-dynamics.com/about/pulleys.htm

The pulley is our design. We had specified the connection to be a butt weld and ground on the inside. This is to remove any geometric discontinuity that are normally present when the shell shape is not in line with the end disk mating diameters. The manufacturer did not grind the inside as specified and did not certify the inspection procedure.

We were not involved in the inspection. The pulley was correctly rebuilt and ran trouble free for many years.

It is not the failure you describe. Your type of failure usually starts at the inner corner and cuts a 45 degree fracture toward the shell and end disk exterior.

L Nordell ■

Dale,

I am assuming from the description of a convex weld on the inside that these pulleys were welded from the outside and the convex weld seen is the back of the root run. It is nigh on impossible to control this weld bead shape when welding from the outside on conveyor pulleys of the size you have indicated (I should know as I have tried for the past 12 years with no success). I have to agree with Mr Nordells comments that a single sided weldment is the worst configuration. The internal weld end is highly irregular and has the highest stress raiser. However, in Australia most conveyor pulley manufactures continue to offer end disc to shell welds on this basis, due to end clients buying the cheapest product and not considering all the technical aspects of a high quality pulley manufacturer.

In your case you have two options as I see it. You can either specify the “T” shape end disc designs as suggested by Mr Nordell or use pulleys that are welded from both sides from a reputable manufacturer that has the ability to finitely control the internal weld bead.

On the design aspect, there is no accurate way to model the weld configuration with FEA. All the FEA will tell you is where the high stress point is but not accurately tell you what the stress level is.

The only way to overcome this stress concentration is as mentioned above accurate internal welding and also minimising any notch effects by additional grinding of this area. ■