The Commando Frame - NOC Roadholder May 2018

Status
Not open for further replies.
I would have thought that a sensible engineering approach would be to build a prototype and test it to destruction, then base your configuration management on it ?

Works for nuclear reactors, right? This is getting silly. The sensible approach is to come from a statistical certainty. Case in point are designs which are within the endurance limit of steel.
 
A statistical certainty is a probability of 1 - it can never reach that unless you arrange for a reactor to self-destruct. Nuclear energy generation can never be risk-free. A definition of safe is 'a situation or condition where the risks are controlled to a level which is tolerable to all stakeholders'. So if your wife won't cop you racing because of the risks - it is 'unsafe'. When you apply that to a reactor, who are the stakeholders ? Sometimes the cost/benefit analysis is unethical.
 
A statistical certainty is a probability of 1
According to you. It can also be a probability of 0. I drive over properly designed bridges with certainty. The nuclear reactor remark was sarcasm and I am certain it was missed by a few.

So what does all this have to do with the wonderfully designed and exceptional (unapproachable)isolastic Commando frame? Lost the plot?
 
Last edited:
IIRC I think I read once that Dr Bauer predicted that Ken Sprayson's idea wouldn't work. I'm glad he wasn't my doctor.
 
I noticed the statement that ideally the gusset plate would have been bigger. It’s clearly seen as an important part of the frame design and function.
I wonder if cutting it away (for fitment of bigger carbs) has any, as yet unseen, detrimental side effects?
 
Wondering the same fast eddie. I’ve only seen a few like that and never heard of a problem. I’ve personally broken one down tube of a widow maker frame (funny little story behind that one) and fractured the verticals of the frame loop just above the little triangular gusset above the rear iso mount. I attributed the latter to general race track abuse and poorly set up isolastics.
 
this video was linked here before , but here it is again. Could not contact the person that posted it on YouTube, as I desperately wanted a good copy of it. So if anyone can get a non compressed copy please let me know.

 
When you ask engineers about safety factors most say 'we think of a number and double it'. For steel that gives a good result in the construction industry. Many engineers don't seem to know that safety factors for materials are usually a percentage of the 0.1% proof stress (the limit of elasticity) and are based on probability. I've worked with many engineers and most have the mindset that you plug in a number and get an answer. The danger lies in them getting answers from scientists' black boxes and never doing an error analysis.

Well ..... so-called safety factors are a product of many single factors, ranging from load uncertainties, scatter of material properties, to manufacturing inaccuracies. There is also a difference between LRFD and ASD design in the way uncertainties are handled. "When you ask engineers ...." - oh yeah, which ones? And those "many engineers" you refer to seem pretty outdated to me! :) There is also a big difference between college graduates with a bachelor degree (who seem to learn up only), and university graduates obtaining a master's degree or a Ph.D. Obviously you have had little experience with mechanical and construction engineers: The proof stress of steel and aluminum is 0.2%.
 
I would have thought that a sensible engineering approach would be to build a prototype and test it to destruction, then base your configuration management on it ?

No. The sensible engineering approach is to sketch up an idea, test it by computer simulation, and THEN build a protype. Unless simulation was erroneous the prototype should hold all static stresses. The dynamic stresses can be simulated to a certain degree, but in the end you have to test it to x cycles (normally 10^8 cycles for a car or a motorbike) to trace the bugs and ensure it's fit for service.

I read somewhere ( I think it was in Classic Bike) that the Commando frame was the first frame ever developed using the finite element method (FEM) and a computer. I missed the mentioning of this in the article above, and I am not convinced it is true. I also missed the information that the backbone spine design and the rubber mounting was a concept inherited from the Z26 project, which by itself was a succession of the ill-fated P10 project at AMC. Now, this is a fact.

-Knut
 
I would have thought that a sensible engineering approach would be to build a prototype and test it to destruction, then base your configuration management on it ? I once worked in a factory where every item in a batch of product was a prototype. We had the same British management culture and hierarchies of workers.
In some applications testing to destrution is mandatory. But still it is a difficult thing to just do on ones own. I have had many assorted presentations made to me by those from this or that prestigious institution proposing that their algorithm has absolute and perfect predictive value. And, as the powerpoint clearly shows it (whatever "it" happens to be) performs flawlessly demonstrating they were worth every penny we paid. And they are all smiles. Then I say, "Now show me where it fails." And how they then become defensive. But all these things from commando frames to algorithms predicting the performance of, well, commando frames, fail. And that is almost the most important part. No one wants to discover catastrophic structural failure in a high speed turn whether it's on the back of a commando or 15,000 ft in the air at 800 mph. The problem seems universal to me. One of the most difficult things for a developer (whether PhD or masters or lone inventer) to do after getting something to work is to willingly and unbidden and describe the limitations of what they've done.
 
I read somewhere ( I think it was in Classic Bike) that the Commando frame was the first frame ever developed using the finite element method (FEM) and a computer. I missed the mentioning of this in the article above, and I am not convinced it is true. I also missed the information that the backbone spine design and the rubber mounting was a concept inherited from the Z26 project, which by itself was a succession of the ill-fated P10 project at AMC. Now, this is a fact.

-Knut

Could be, but I kind of doubt it. Although there are limited earlier examples, regular use of finite element analysis (FEA) for such structures didn't really pick up until the late '60s and early '70s, when some good programs (NASA made NASTRAN publicly available somewhere near 1970) were released. And even then, it really wasn't much used unless you had access to a really powerful computer. By the late '80s, or maybe even early '90s, it was developed to the point where FEA could be done on a really good desktop computer, and lots of commercial software was available. Back in the '60s, we were still inputing data on punched card decks in a batch process on remote mainframe computers. My application was FEM computation of electromagnetic fields, and that wasn't really practical outside Universities, Governments, or major Companies (Raytheon, Hughes, etc.) until maybe the late '80s. If Norton used it in designing the Commando frame, it had to have been in conjunction with one of the Universities. I'm making these timeframe estimates from memory, and certainly could be off a little, but I'm quite sure Norton didn't have that sort of computational power in house in the late '60s.

Now, anyone with some engineering knowledge, and spare cash can buy one of many sophisticated FEA packages for a few thousand dollars, and a desktop computer for a similar amount, and be in business. Technology moves pretty fast nowdays.

Ken
 
Another reason to doubt any FEM analysis is that if they had done one, it would have shown a huge stress concentration under braking loads at the point where the early "widowmaker" frames broke. No competent design engineer could have missed that.

Ken
 
Another reason to doubt any FEM analysis is that if they had done one, it would have shown a huge stress concentration under braking loads at the point where the early "widowmaker" frames broke. No competent design engineer could have missed that.

Ken

Well dont forget he came from Rolls -Royce and might have had access to facilities there ,possibly on a personal and very limited basis Their computing capacity would have been stretched with rB211 associated problems
 
Another reason to doubt any FEM analysis is that if they had done one, it would have shown a huge stress concentration under braking loads at the point where the early "widowmaker" frames broke. No competent design engineer could have missed that.

Not necessarily. Stress concentrations become evident only if the mesh is refined adequately, and of course if the load case caters for this particular load distribution. The stress concentration may have shown up as not sufficiently high to raise the flag. A further mesh refinement would have shown it. With limited software capabilities (meshing would have been performed manually) and very expensive computer time, in-depth analysis of the headstock joint may have been impossible.

-Knut
 
Could be, but I kind of doubt it. Although there are limited earlier examples, regular use of finite element analysis (FEA) for such structures didn't really pick up until the late '60s and early '70s, when some good programs (NASA made NASTRAN publicly available somewhere near 1970) were released. And even then, it really wasn't much used unless you had access to a really powerful computer. By the late '80s, or maybe even early '90s, it was developed to the point where FEA could be done on a really good desktop computer, and lots of commercial software was available. Back in the '60s, we were still inputing data on punched card decks in a batch process on remote mainframe computers. My application was FEM computation of electromagnetic fields, and that wasn't really practical outside Universities, Governments, or major Companies (Raytheon, Hughes, etc.) until maybe the late '80s. If Norton used it in designing the Commando frame, it had to have been in conjunction with one of the Universities. I'm making these timeframe estimates from memory, and certainly could be off a little, but I'm quite sure Norton didn't have that sort of computational power in house in the late '60s.

The NASTRAN code developed by CSC (Computer Science Corporation, Virgina, USA) was based on code fragments already developed by NASA since the early 60's. The complete code was released to NASA during 1968. Interestingly, work at CSC has ties to the Cambridge Engineering Laboratory (UK). At the same time the development of the SAP code was on high gear at Berkely, California. Eventually, SAP V1 was realeased in September 1970. It is not unthinkable that one of the developers of either code was charged with performing limited analyses for N-V during 1967.

Anyhow, the prerequisite of the FEM method is matrix algebra. In 1964 the book "Matrix Methods in Structural Analysis" by R. K. Livesley was published along with code for frame analysis, which spurred widespread interest in using automated algebra to simplify structural analysis. R.K. Livesley was with the Computing Laboratory at Manchester University since 1951, advancing the use of computers in structural mechanics. At the time, computers had to be programmed in machine code which was a daunting task.
Livesley went on to work at Churchill College, Cambridge, UK. With the introduction of general purpose computer programming languages like Algol (1958) , Fortran II (1958) and Fortran 66 (1966), a paved road for harnessing computing power in structural analysis existed.

I am sure N-V were aware of these developments, and certainly R-R. By 1967 matrix algebra in structural analysis was well established in construction and mechanical engineering and it would have been rather easy for N-V to apply this technology, and it's likely they did this, at least.

-Knut
 
Last edited:
I seriously doubt that any kind of CAD was involved, unless someone had access to a university computer. N-V didn't have one. It didn't have any test instrumentation at Wolverhampton either. That's what I was hired for, since I had a lot of background with British Aircraft Corporation flight testing. I was disappointed to find that the budget allocated for the development of an instrumentation system had been spent with the idiot PR company that persuaded N-V to abandon the historic "Norton" logo and go with that ridiculous "green globe" insignia. N-V continued with the "rider opinion" method of testing and I became one of the road-test riders.

A wasted opportunity to get the company into a more sound engineering way of doing things, and a big disappointment to me, hence my emigration to go work for Boeing in mid-1968.
 
what's the story (why?) with going from rr to norton?

big money involved?

falling out with rr?

lured away?
 
Good question... R.R were then at the pinnacle of jet engine technology . Heard it was down to Dennis Poore who somehow knew him...
 
Last edited:
It may have been a combination of having reached a plateau at Rolls-Royce and having the additional "prestige" of actually being a Managing Director. He was a decent guy, but a bit aloof from the "grunts". I dealt with him a bit more than the other grunts could when I did the high-speed cinematography on the AJS moto-cross bikes.

I remember his surprised gasp when he saw the swing arm of the bike go from fully down (bike airborne) to hitting the rider in the backside (arm still fully down) after a jump. It took just two movie frames (about .03 seconds). The frame was so stiff that the suspension had to do a lot more work. It only took one race to get the rear dampers so hot, the paint burned off. If the next race was not far enough away, we'd douse them with water (so we didn't get burned), then replace them.

The other engineers at Wolverhampton were still up to their armpits in alligators with the frame stiffness vs. suspension "simplicity" issue on the M-X when I bailed out to emigrate. The production "Stormer" had just hit the market and seemed to be doing well. I almost bought one to bring with me from England, but decided we couldn't afford it.
 
My working life began in 1958. Back then we built on what had gone on previously. Seeley frames were not rocket science. But if someone was elitist and arrogant, they might not want to learn from the experience of others. The Indians are smarter with their Enfields - they seem to be using Harris frames. I cannot comprehend the mindset of an engineer who believes they can design from scratch, something better in motorcycling than what has already performed well in practice. Personally I would start with something which actually performed and develop it further. An Egli Commando might have been a good thing ?
 
Status
Not open for further replies.
Back
Top