Who knows a lot about metallurgy?

[llile]

I have built several shop-made recumbent bikes, and had the same failure over and over despite my best efforts to prevent it.

The "pedal boom" on a 'Bent is the part that sticks out in front of the bike, with the pedals on it. This part takes a lot of strain - repetitive flexing. I am a strong rider, despite being a skinny, scrawny little fellow with a pot belly I can leg press all the weights on the leg machine at the gym. I would wager many bicyclists could do the same.

Over the last two recumbent frames I have built, the pedal boom has fatigued and failed a total of 5 times. Each time, I reinforced the part that failed, only to find it fatiguing and breaking somewhere else. The steel I am using is plenty thick enough to take the strain initially, although it flexes on an uphill start, I couldn't possibly break it, but it finally gives way. I am using chrome moly steel for the pedal boom.

Quite often the failure is right next to a weld. Not on the weld - that is plenty strong, but in the metal next to the weld. This time it was in a flat area that looks like it got hot when a nearby weld was made - an inch from the actual weld. A friend was trying to explain to me some considerations about heat stress in chrome moly from MIG welding. He said some frame builders will heat treat the whole frame, raising it to some temperature then covering it with blankets and letting it cool slowly. I don't have any idea if I can do this in a home shop, or if he had any idea what he was talking about.

But right now I'd like to understand a lot more about metal and heat stress. What is going on next to a MIG weld, stress-wise? How could one make it less prone to failure? Has the heat weakened the steel?

[bigmoose]

I am not an expert, but I think I can give you a few buzz words to google. Look up heat effected zone or HAZ. It is the area near the weld that has weird as in bad properties. What you buddy is recommending is called re-normalizing I think. That heat it up and slow cool removes or highly reduces fabrication stresses.

[liveforphysics]

Up the tubing diameter by a 1/4".

Problem will be solved.

[psycholist]

Look up hydrogen embrittlement. Brazing is the most desirable method for welding chrome-moly.

[llile]

Up the tubing diameter by a 1/4".

Problem will be solved.

Nice try. I also thought of the "big tube theory", after I thought of the Big Hammer Theory (my favorite bluegrass band name..) That broke too, although slower.

In one case I put a nice big fat reinforcing patch on the area where the most leverage occurred. 'That'll fix it!". The patch held, but the tube broke at the weld again, ahead of the patch. I cut the darn thing off and used a bigger tube. It broke too, eventually, at the weld. Pedal booms are often welded onto a bike head tube. One time the head tube split, right next to the weld again. This is the 5th time, I have thicker tubes yet, and a patch on both sides to reinforce it. The break occurred at an area where I welded something onto it. It has to do with heat, I am suspecting, and some other folks are indicating that is the case too. I'm just beginning to understand all this, though, despite having learned to weld a long time ago.

[llile]

Look up hydrogen embrittlement. Brazing is the most desirable method for welding chrome-moly.

I know how to braze, but yow! I gave it up years ago because it requires super-precise joints. I made a half a recumbent frame with brazing one time, over several months, then I bought a MIG welder and finished it in a day. With a weld you can fill in a hole, with brazing the parts have to *fit*. Now, there are some cases where I can make that happen, some not.

That being said, maybe there are specific parts of the pedal boom, where all the high stresses are, that I could braze...... I've never had any other parts break off, that couldn't be traced to a crappy weld or poor weld prep.

[llile]

I am reading some stuff about preheating MIG welds, or heat treating them afterwords with a torch. (Also reading that I shoulda bought a TIG welder... but that isn't in the budget). Don't understand it yet though.

[melodious]

I'm far from an expert, but I have looked at how Catrikes build their aluminum frames. I'm guessing the pedal booms are easily replaced as it's separate from the main frame body. Would putting some kind of "lattice pattern" within the tube give you more rigidity/durability and less flex/potential failure?

[John in CR]

Sounds like bogus 4130 to me, and/or you got 4140, which does require heat treatment. Get real stuff from an aircraft supply and use that larger diameter and double the wall thickness and be done with it.

If you already have the tubing, then shape the joint differently and/or use a reinforcing sleeve cut at an angle or V to spread the stress and heat related weakness along the axis of the boom instead of perpendicular to it.

[Farfle]

Sounds like bogus 4130 to me, and/or you got 4140, which does require heat treatment. Get real stuff from an aircraft supply and use that larger diameter and double the wall thickness and be done with it.

If you already have the tubing, then shape the joint differently and/or use a reinforcing sleeve cut at an angle or V to spread the stress and heat related weakness along the axis of the boom instead of perpendicular to it.

I agree with everything here, if you want to get really fancy, take a rosebud tip on an oxy touch and heat the weld and surrounding area (HAZ) till its dull red, then let cool slowly by slowly backing off the torch.

[Xanda2260]

The problem with heat stress occurs because the weld metal is bloody hot (obviously), once the weld is complete the weld metal contracts as it cools, pulling on the surrounding metal with some fairly epic force as it does so. i'm guessing that preheating the steal would help as all the metal would be cooling and shrinking by similar amounts.

Coming at it from a bifferent angle, do you tend to ride in high gear at low speed a lot?
If so, this will put a huge amount of stress on the joints (and your knees).

[d8veh]

if you're not bothered by a little extra weight, use slightly thicker mild steel tube, which is not as strong - hence the extra thickness- but it doesn't suffer from metal fatigue like CroMo, and it can be welded without causing the same level of embrittlement as CroMo. Othewise it's CroMo and brazing.

[spinningmagnets]

Sometimes a simple gusset can spread the stress out enough over a broader area so that the metal in the frame acts as spring, and the flex matches the frames capabilities. Thats why the lugs for old-style frames had tapered and sometimes ornate butt-ends, to spread out the stresses...

[mat h physics]

The comments here are in the right ball park, but I think you problem is compound. Cracks are starting due to: A) the metal is too hard B) The fillet/gusset is "rough" causing stress fracturing.

Solution: Smooth the weld area by sanding, esp where it mates the parent metal. Grinding will work, but may create a place for the stress fracture to start, paper will flex to help prevent this. Welds have to be annealed, esp in complex alloys. I don't see how blankets can hold enough heat to do this, but cooling off slowly is the key. Using a propane torch (or mapp gas for a hotter flame, find recommended temp for parent metal) heat the weld area and keep running the flame back & forth, allowing it to cool slowly (gradually backing away). This process will take more than a half hour, and that is fast, it should be done in an oven over several hours.

The weld hardens the metal to a glass like state. For a dramatic understanding of this, look into glass blowing. The glass will crack if not annealed (90% of the time). Any rough texture only increases the chances a crack will start.

Trivia: When transporting liquid O2 through pipes, the interior has to be smooth as a babies bottom. Even a sharp edge left on an elbow will create eddy currents w/ enough energy to ignite the oxygen in a liquid state (even radio waves have been enough).

[psycholist]

Lots of good advice here. Another problem not mentioned yet is the resulting undercut at the edge of an electric weld where it meets the parent metal. This is another reason why brazing is preferable. Seems your a strong pedaler and since this tube is only joined on one end, the constant oscillating leverage exerted on this brittle, hard, thin-walled joint is causing it to fatigue and eventually fail. I favour d8veh's solution using thicker mild steel and reinforcing the joint with a sleeve or gussets. If possible, weld a triangular gusset, cut from flatbar, along the length of the tube and crossing over the joint. One on each side, if you can manage to position it so it doesn't interfere with pedaling.

[llile]

Sounds like bogus 4130 to me, and/or you got 4140, which does require heat treatment. Get real stuff from an aircraft supply and use that larger diameter and double the wall thickness and be done with it.

If you already have the tubing, then shape the joint differently and/or use a reinforcing sleeve cut at an angle or V to spread the stress and heat related weakness along the axis of the boom instead of perpendicular to it.

Well, i *do* buy my 4130 from an aircraft supply house. The "V" cut reinforcing sleeve is a good idea - I have used perpendicular cut reinforcing sleeves before, and they just moved the stress around to somewhere else. The more I read about MIG welding 4130, and I am just beginning to understand this, the more I begin to think that I either need to preheat the joint or heat treat it afterward. Experts on welding forums either say "Use TIG on 4130, forget MIG" or "Yeah, you can use MIG on 4130 if you do this and that" However, the exact recipe for this and that is different every time I read about it.

Yes, I have gone to lengths to add reinforcement to joints. One of the joints (that failed, eventually) had a short internal tube welded onto the head tube, then the pedal boom was welded over that, so it functioned like the old cast lugs on frames. I can't remember if that one failed at the head tube, or just ahead of the reinforcement.

This has only been an issue with joints that are taking a lot of tension and repeated flex. The rest of the frame has always held together if they were a part of a triangulated assembly. I have a part on my new project that will be in the same kind of stress - which is reason #2 I am trying to learn more about this issue.

[llile]

Here are some photos of the result. I ground out the crack int eh tube, and welded it back with a MIG, then ground it flat. Then I brazed a piece of mild steel angle iron the length of the tube. I was careful to heat the whole tube, and the welds on it, and let them air cool as slowly as possible. Sure, it isn't like putting a blanket over it, but I had no asbestos blankets handy. I will have no idea if it worked for a year, if I am still riding this bike a year from now. (That will mean my next bike has been a complete failure...)

P1020956.JPG (200.12 KiB) Viewed 218 times

P1020957.JPG (208.6 KiB) Viewed 218 times

[Kirk]

If you can find one around. Take a look at the cup gusset underneath a Catrike's boom socket just in front of the cruciform. Also the way Paulo cured his booms problems with flexing and cracking was to add a three point reinforcing web inside of the boom. Maybe you could just order one their's and cut it to length, might be simpler that way.