When is a DIY frame safe?
Post Reply    Forum -> Not-So-Modern
Author Message
Addicted
2006 LX150 (carbed) | 2007 GT200
Joined: 29 Jun 2016
Posts: 797
Location: Toronto
Fri Feb 07, 2020 7:16 am quote
My question is inspired by charlieman's (awesome!) thread, but is not meant as judgment or veiled criticism. Just an honest curious question: For those of you that do safety-critical major fabrication like frame building, how do you get comfortable that it's safe?

Presumably a one-off home project doesn't get dynamic stress modelling; engineering review; destructive testing; weld inspection; etc, etc. Is it experience based on doing similar work professionally in day jobs? Comparison to other manufacturer projects? Road testing after it's built?
Banned
2:6
Joined: 11 Jan 2007
Posts: 7507
Location: San Francisco
Fri Feb 07, 2020 9:05 am quote
Most of the time when modifying existing stuff I just go for overkill on strength for the benefit on stiffness and do it based on "engineering judgment".

but on fully new stuff I'll model the parts and add the loads in for an FEA. assume 5x for gforce, 2x for step loading, 1/2 the yield strength for the material to account for fatigue strength, and then 2x safety factor (higher if the loads are in question).
Molto Verboso
1963 VBB2T
Joined: 07 Nov 2012
Posts: 1376

Fri Feb 07, 2020 9:20 am quote
oopsclunkthud wrote:
Most of the time when modifying existing stuff I just go for overkill on strength for the benefit on stiffness and do it based on "engineering judgment".

but on fully new stuff I'll model the parts and add the loads in for an FEA. assume 5x for gforce, 2x for step loading, 1/2 the yield strength for the material to account for fatigue strength, and then 2x safety factor (higher if the loads are in question).
Yep this procedure should cover it. Personally I hate math and just weld for good penetration and then cross my fingers for luck. But in all seriousness a good weld and a good gestimate on where the weakness of a project may appear will get you pretty darn far.
Veni, Vidi, Posti
74 Super, 75 Super, PX project, LML off-roader and new to 2018, '66 Blue Badge Smallframe
Joined: 30 Nov 2011
Posts: 7276
Location: Victoria, Australia
Fri Feb 07, 2020 6:38 pm quote
How about: see where other similar things have issues/failures, and improve them? Real-world destructive testing.
Ossessionato
Joined: 26 Oct 2015
Posts: 2080

Fri Feb 07, 2020 8:18 pm quote
Itís much easier nowadays with the internet. It used to be that I would go down to the local steel fab shop, and they would look in their books to find out the ratings of the steel. Iíve welded up a few railing and hand rails for both residential and commercial businesses. The engineers required specific wall thickness for the metal. Also having worked in the automotive repair industry, there is a fair amount of structural sheet metal work. Iíve also built a couple link suspensions for my jeeps. After years of working with metal, a person will just know what is both safe and strong (and yet not too heavy). I donít know Charlieís background, but if he used to mess around with old tuned up Dastuns, heís probably versed with whatís required for fab work...even if he does or doesnít have first hand experience building stuff, it looks like what heís building is spot on to me. I havenít seen any red flags (or even yellow flags) on any unsafe things on his hack. Google searching will get you a looong way nowadays for information. Caster, camber, scrub, toe-in/out, rake, trail, travel, stroke, spring rates, etc..
Ossessionato
2015 GTS300, 1974 Primavera, 04 Ninja 250
Joined: 04 Apr 2013
Posts: 4286
Location: San Diego, CA
Sat Feb 08, 2020 6:59 am quote
Which thread inspired the question?
Addicted
Joined: 26 Jan 2019
Posts: 871
Location: california
Sat Feb 08, 2020 7:31 am quote
Quote:
not meant as judgment or veiled criticism. Just an honest curious question: For those of you that do safety-critical major fabrication like frame building, how do you get comfortable that it's safe?
Berto - thanks.
Took it as a compliment
While it may look like wild guessing - the boil down is.
"see where other similar things have issues/failures, and improve them"
"Real-world destructive testing."
To that - I would add - practical experience - from trying other smaller stuff.
Lots of examples in my current build

It may seam complex/risky - but complex stuff is just a bunch of simple steps put together in a string. My thread is full of me learning a bunch of simpler steps - and learning from this forum and others. Watching LynnB cut out and re-weld a floor board for example - simplified some of the subtle steps needed for the sheet metal welding portion. I also chose my tube size and type after looking up what motorcycles use for their frame as reference. I'm sure there are 100 more examples that have lead to my understanding.

When I am done - I will test carefully.
If it doesn't work - I will tear it down and alter it where needed.
In fact - I will be doing that today.
Other leaner sidecar projects from the internet have taught me what to look for that will make it ride poorly - and my current build is full of them!
The simple mock up I made allowed me to test and learn.
Now I will iterate as Ginch suggests:
Quote:
see where other similar things have issues/failures, and improve them
There is probably less guessing then it might appear...
-CM
Addicted
2006 LX150 (carbed) | 2007 GT200
Joined: 29 Jun 2016
Posts: 797
Location: Toronto
Sun Feb 09, 2020 7:20 am quote
Thanks to all for the feedback!

I'm personally more in the oopsclunkthud camp of running FEA, but I think building based on similar examples would be fine too. It's actually my lack of fabrication skill (in particular, welding) that would scare me most. But I can certainly understand other approaches and different comfort levels.

For the record, I don't think anything I've seen would count as "wild guessing"! I am always really impressed by the projects on this forum!
Banned
2:6
Joined: 11 Jan 2007
Posts: 7507
Location: San Francisco
Sun Feb 09, 2020 7:56 am quote
short of FEA but beyond just experience is method were you look at the stress at a few points in the structure. yes you have to run the math, but the art is to pick the points where the stress would be highest. If they are good, the rest should be fine.
Hooked
piaggio hexagon
Joined: 17 Aug 2013
Posts: 449
Location: lincolnshire uk
Sun Feb 09, 2020 11:36 am quote
I once read a (probably apocryphal) story concerning the differing design philosophies of German WW2 aircraft designers Willi Messerschmitt, (Bf109) and Kurt Tank (FW190). It was said that if a part failed during testing on a FW190, Tank would strengthen it, if a part failed on a Bf109, Messerschmitt lightened the structure around it!
Addicted
2006 LX150 (carbed) | 2007 GT200
Joined: 29 Jun 2016
Posts: 797
Location: Toronto
Sun Feb 09, 2020 3:06 pm quote
oopsclunkthud wrote:
short of FEA but beyond just experience is method were you look at the stress at a few points in the structure. yes you have to run the math, but the art is to pick the points where the stress would be highest. If they are good, the rest should be fine.
Makes sense. And the large factors of safety I suppose should cover for manufacturing variance from theoretical / ideal case.
Rallies Europe 2016   Vespa Wasp Pin Badges   AF1 Racing Vespa Austin
Post Reply    Forum -> Not-So-Modern
[ Time: 0.1265s ][ Queries: 25 (0.0350s) ][ Debug on ]