Metco idler broke. Pics inside

[Silver_Serpent]

Tob ... it looks like you applied a side load of 100lbs? I dont really know what the designed belt tension in these cars is, but if you applied the reaction force of the belt tension on the pulley instead of the side load, you should get the additive effect of the tension and bending moment that simulates the actual loading. Nice work!

 

[bFoxGt500]

I couldn't remember where I got my tensioner when I posted earlier. It is a Thump. High quality piece. FWIW.

 

[mnewxcv]

after seeing the stress tests, its interesting to see. Question, is there a certain number thats considered the breaking point for that material, and what is it? as in, I see it peaks at 21,000 in the last stress test.

 

[Silver_Serpent]

after seeing the stress tests, its interesting to see. Question, is there a certain number thats considered the breaking point for that material, and what is it? as in, I see it peaks at 21,000 in the last stress test.

40,000 psi is the point at which the material yields. However, for this application, I would believe that 21,000 psi would be too much for this aluminum that will experience so many heat cycles. But, I think Tob has kind of just "thrown some loading on it" for now to see the stress patterns. I would bet a side load of 100 lbs is a little overkill compared to what the real side loading is based on belt tension.
With that large of a load, and still only being at 50% of material yield, is actually pretty encouraging for the piece (I would have guessed 100 lbs to the side of the pulley could have bent the thing).

 

[Tob]

Tob ... it looks like you applied a side load of 100lbs? I dont really know what the designed belt tension in these cars is, but if you applied the reaction force of the belt tension on the pulley instead of the side load, you should get the additive effect of the tension and bending moment that simulates the actual loading. Nice work!

It was late and I just wanted to see if the simulation software I have in Solidworks was going to give me any flak. It did. First, I can't run the simulation on a parts assembly so I combined the major elements (pulley, main bracket, and standoff's) into one part. Which is fine, as the pulley is bolted to the bracket and the bracket is fixed to the engine. I then chose 6061-T6 as the material to be stressed as that is what Metco advertises it as being. I couldn't apply a direct force to simulate belt tension so I substituted as you noted. I would like to test it that way and will take a deeper look as soon as I have some time. I was going to sketch a force diagram to estimate the hubload angles from a number of photos I've found that should be fairly representative. I'm also trying to formulate a typical load that a typical GT500 belt imparts on a given pulley as well. I looked again at one of my favorite SAE papers since it was done by Roush/Ford on the Ford GT front end accessory drive for some numbers but they weren't included (scroll down to about figure 15 and you'll see what I'm referring to).

40,000 psi is the point at which the material yields. However, for this application, I would believe that 21,000 psi would be too much for this aluminum that will experience so many heat cycles. But, I think Tob has kind of just "thrown some loading on it" for now to see the stress patterns. I would bet a side load of 100 lbs is a little overkill compared to what the real side loading is based on belt tension.
With that large of a load, and still only being at 50% of material yield, is actually pretty encouraging for the piece (I would have guessed 100 lbs to the side of the pulley could have bent the thing).

You are pretty damn perceptive! It was late and I just wanted to feed in some numbers to see if the impending virtual failure would mimic the dynamic real-world failure the OP suffered on his car. I'll try a plethora of load values and see how low I can go and not see appreciable damage to the bracket. I'm also going to substitute steel(s) for aluminum and see how much better they do. This has been an interesting exercise nonetheless.

And Serpent, if you've got some good simulation software, I can forward you the assembly I combined if you'd like to analyze it as well. By all means let me know if you do.

 

[BROK3N-SVT]

I seen this video of a zl1 tensioner during acceleration. I just remembered it and hunted it down. This might give an idea of what kind of force is being applied under load. It's kinda cool.

http://youtu.be/5NnMuyifPA4

 

[Robert M]

So who else has had issues with a Metco idler? I have seen many of these installed over the past few years, how many have failed? I don't remember reading about Metco idler failures as being the norm on any of the forums that I follow? If this is a common problem, I would guess that there are more who have had issues.

Maybe I do remember someone having issues with bearings a couple of years ago...............



R

 

[BROK3N-SVT]

I have never heard of it happening and that's why I posted it. I know bearings fail often and mine feels fine but I ordered a new one the other day. Might as well for only $3.50 each.

 

[03 DSG Snake]

Almost there, back when the three year old design was new.

I've been running the Steeda setup for awhile, after ditching my Metco. I had problems with chirping belts, even after swapping to the double bearing Metco idler. Added a Thump tensioner in between, after swapping to my 3rd belt the chirps went away. The belt loves to dust however. VMP thinks my Reichard pulley is grooved incorrectly. I plan to fix it all at once with a TVS swap when I get a chance.

The Steeda is a pretty solid unit, however. The idler has a nasty tendency to turn belts silver though (looks like they have a different idler now).

 

[03 DSG Snake]

Updated Metco piece:

 

[SuperFly03]

That seems much stronger than the OP's bracket, which is what I unfortunately have.

 

[Tob]

Shelby version aside, I've seen four different variations from Metco. The oldest photo I could find was circa '08 or so and I show it at the top in the following photo sequence. The basic shape has remained throughout each following iteration but it differs from subsequent designs in that it was a solid slab with no triangular sections of material removed. I then noticed a thin followed by a thick(er) version. And now it looks as if the triangular windows have been closed, albeit with reduced thickness. Certainly an interesting "progression" if I'm following this correctly. They'll eventually get back to where they started.

 

[Robert M]

I think, but am not positive that mine is the thicker frame with the cut outs, at least it seemed to look thicker than some I have seen......



R

 

[BROK3N-SVT]

Well just got the replacement in the mail. He just sent me the whole thing. It is bad ass and looks very strong. Rick from metco took great care of me.

 

[mnewxcv]

nice! side by side with the old?

 

[Robert M]

Sounds like Metco has Great customer service!!



R

 

[Silver_Serpent]

Deffinetly have to appreciate companies that take care of their customers, no questions asked. We are fortunate to have several vendors like that with this car.

 

[Tob]

It does say a lot about Metco. Good to hear they are taking care of you.

Before this one fades off into the abyss, I wanted to add a few things. I got caught up breezing through a number of SAE papers on belts, belt tension/pulley design, frequencies/harmonics and excitation. The whole accessory drive system has quite a bit of engineering behind it. Small changes can have a major impact on the system's overall effectiveness. I honestly never paid that much attention to the system other than routine maintenance prior to this. Knowing what I know now, I'm even more impressed with the latest belts from Gates that Van is carrying for this system.

Anyway, one of the things I was looking for was appropriate belt tension, more specifically, that of the blower belt. I searched through Ford service manuals and a few other resources but didn't find much beyond a recommended check of the factory belt tensioner at its range indicator. I did find this chart while perusing Gates' Master Catalog (note the highlighted line for the 10 rib such as the blower belt)...

I had no idea the numbers were that high, especially those for a newly installed belt (far right column). At least I had some data to plug in or at least a range for some additional simulation. I also wanted to see how the angular vector force changes with the addition of another pulley such as that of the Metco idler being discussed in this thread. So I cleaned up a sketch I found in a service manual that illustrates the blower belt on our beloved GT500's...

Obviously, that's the blower/pulley at the highest point of the sketch. Note the span between the blower pulley as it travels down to the left and makes contact with a "smooth fixed idler" pulley that all of our cars came equipped with. Now take a look at the same sketch with the addition of an auxiliary idler between the blower pulley and the smooth idler...

The addition of the auxiliary idler reduces the wrap on the factory smooth idler which has an effect on the frequency order between the two pulleys. I depicted the auxiliary idler as accurately as I could relative to how it would be positioned factory GT500 5.4 engine. The vector at which the tension pulls is through the center of the wrap it has on the pulley or up and to the left. This was the angle I was trying to use when directing a force or pressure while trying to analyze the Metco bracket. I was using Solidworks SimulationXpress for the analysis and it wouldn't allow me to test an assembly (pulley, bracket, standoff's, etc). So I combined those elements into one part and replaced the circular pulley with a square block that was placed at a specific angle that would be representative of the vector force from the Metco pulley (still with me?). I then ran a number of analyses, with either pressure or force aimed squarely at the block face. I varied the numbers used each time and then ran the tests over with the bracket done in steel instead of 6061-T6 aluminum. I did run some simulation through Solidworks Simulation (which goes into far greater detail than Xpress) after some time.

The results were very similar to the quick and rather arbitrary tests I posted up on earlier. The bracket is stressed most in the same area, precisely where the OP's failure occurred. I have to say, Solidworks impressed the living heck out of me in this regard with its simulation software. It even generates multipage analysis reports on the matter if you ask it to...

I'm actually thankful for the opportunity this failure gave me to take a deeper look into modeling and testing. I learned a lot and will apply that knowledge towards a few other projects I'm working on. I wanted to do more work on an optimized bracket but at this point I'm moving on. But Ron, just for you, I rendered this piece in steel plate with a Lipstick Red powdercoating since I know how much you like pizzazz...

This was a truly interesting thread. Thanks OP!!

 

[Van@RevanRacing]

Thanks Tob. Now my brain hurts. Nice thorough analysis.

 

[Tob]

I know, mine too, and I greatly simplified the above for this application. But if you are the type that enjoys a good technical read I give to you pain, pain, and more pain.

https://www.youtube.com/watch?v=v9AHIwmH7AY