Fatigue Testing - Main Tube Gusset Comparision

I am planning to run a series of fatigue life tests and compare different gusset styles, 2 filler materials, and maybe a brazed joint. Tubing will be 35mm (1-3/8") because that’s what the current fixture is setup for. Loading scaled to simulate the ISO 1,200N horizontal fatigue test. Tubing will be mostly NOVA Crmo because it’s cheap. I’ll do a couple tests with Columbus Zona also for comparison, and consider expanding to some other materials depending on interest and findings. I’ll post the proposed gusset designs here for discussion prior to making them so we can debate the merits and determine what is the most interesting style to test.

Goal: generate some experimental numbers to compare gusset options and tube materials.

Here’s a shot of the machine. I’ll see if I can figure out a way to share some videos (there are a couple in my Instagram already)

More to come…


That’s great! I’d love to see the brazed joint for sure.

1 Like

This is great! Do you have more details about your data acquisition and actuator for the loading? I have been wanting to do this kinda testing forever.



1 Like

@Daniel_Y I’m using a Dataq DI-2108 for data acquisition:

This included load cell, linear displacement, supply pressure, and actuator cylinder pressure.

The machine uses pneumatic cylinders for loading is controlled by an Automation Direct PLC. Let me know if you want more details on sizing and models.


@drwelby No, I have not seen that report. Something you can share or direct me to a source? Thank you!

Very cool. Excited to see what type of fracture mode comes out of this. Have made some effort into understanding “at the toe” fractures, but wasn’t able to replicate it with the tensile testing machines at school so was wondering if a fatigue testing setup similar to this would. I’m guessing the amount of strain per cycle is critical.

I can’t for the life of me remember where the link to this is, but I found Bob Brown’s analysis that includes Bontrager’s analysis:

1 Like

After you do a few of these, you start to understand why a high speed machine is so desirable. I did my thesis on a 10Hz machine and then later bought one that would go 100hz. Still suffered from grad students not liking how slow the tests went.

I would love to see BMX style plate gussets tested, alongside the sort of mondraker/BTR style (tube from HT-DT junction to TT).

For no self-serving reason what so ever… yeah…




That was fascinating. Pretty rigorous stuff for an enthusiast mag, although I may be mischaracterising it as it predates me.

1 Like

This test setup is pretty rad! Those Dataqs are awesome for the price.

These are both on my list

Any updates on testing?

I’ve been consumed with other stuff, so no update other than the sample tubes are here. I picked up a titanium tube also just for fun to see how it compares.

I’m also looking at updating the test fixture to improve repeatability, but haven’t figured out a design that I like yet that doesn’t involve lots of expensive components.


Not much of an update, but an update all the same:
I’ve had a little rime to think through a revised design for the saddle that interfaces with the tube. Still have to finalize and order parts.

I picked up a titanium headtube and toptube to add to the list of test subjects. Should be interesting.


Some progress today…
Mom is visiting so I put here to work mitering tubes on the mill!

Headtubes are also in process.

Redesign of the tube “saddle” for the fatigue machine is almost complete, then parts can be ordered.

It will be time to discuss what everyone would like to see tested next!


Parts are here for the tube saddle, but it will be a bit before there’s time to build it. Have some other tests to run first and a couple frames to finish.


Curious if I could send you two different joints for testing at some point.

Your set up is for 1-3/8 tubes? 44mm ht?