Engineering disasters and mysteries.

[2grimjim]

I don't really have any specific examples but I would like to hear opinions and first hand experience of after affects of bone-headed engineering, at least how it applies to motorcycles.

Not that I'm picking on anyone with an engineering degree, but I've had to work with engineers with Masters and Doctorate degrees in various disciplines of engineering, and they might be smart, but sometimes the simplest things escape their grasp. I also understand that a lot of mechanical engineering is done by accountants, thermal dynamacists, electrical engineers, and the occational engineer with a forrestry degree.

That being said, I'll confess that I do not have an engineering degree, but I don't think that makes me unqualified to pick on them anyway. My father retired as the head of a mechanisms design group for a large Aerospace manufacturing company after 39+ years in the industry. He did not have any formal education past high school, however, every one that worked under him did have a university degree of some sort. His biggest complaint with the staff that worked for him was that, when he had a new engineer start working for him, the first thing he had to do was make them spend a month on the manufacturing floor to "figure out how things are really made".

Just to start things off (oh, I'm sure I will piss off all of the Honda guys) I'll throw out the horrible valvetrain design of the '02-'08 CRF450R/X. The first time I worked on one of these was in '02. It wouldn't start (a soon to be familiar theme with these bikes). When I poped the valve cover off I was mortified that the entire valve train was held to the head with four 6mm bolts that were very long (hense 'very strechy'). The extra long 6mm bolts are more like 6mm springs, allowing the cam assembly bounce and hop on the cylinder head. The other shocking observation was the suprisingly small (narrow) cam bearings used. After doing a few of these top ends I discovered that the ones that were the chronic valve eaters all had one thing in common; the cam bearings were worn enough that I could get up to .003" radial movement in a ball bearing. It was usually the cam sprocket side, the one that is not replacable. The effect is obvious; you can't posibly set the valve clearance correctly with worn out cam bearings.

Honda did change this design in '09 and fixed both items. The cam carrier was incorporated into the head casting (doing away with the spring/bolts) and they increased the size of the cam bearings. Now the CRF is as reliable as any other four stoke MX bike.

Edited January 4, 2012 by 2grimjim

[KJ790]

As someone who recently completed a masters in mechanical engineering and has a job as an engineer, I completely understand how disasters happen. I don't get along with many engineers because they don't seem to have a grasp of real world application. I come from a family of mechanics, and worked as one before college and through a few of my college years. I then worked as a composites fabricator for a year in college, and then did 2 years as a research engineer before getting my current job as an engineer at a metal finishing plant.

My last year in school I discovered that most of the mechanical engineers that were about to graduate did not know how simple mechanical devices worked. I was shocked when my senior design class couldn't figure out how to make a shaft rotate in one direction and not the other. I suggested to make a ratchet and I then had to explain how a ratchet works... One kid then said "Wow, did you just come up with that? That's really smart!" ?

Engineering schools tend to focus on theory and never touch on actual application or the nuts and bolts of how things are held together. Simple things like fasteners are never even brought up, so most new engineers never bother to think about how things are actually going to be held together. If a person doesn't take it upon himself to gain real world application experience and application of theory, he will never know it and will continue on in life making poor designs.

Good engineers are the ones that bother to learn how things actually work, not just the theory behind how they should work.

That being said, there are so many things that come into play in real world application that not everything can be calculated or accounted for, especially when there is a timeline to get a product finished. Add in the fact that the bean counters are the ones that get final say for designs and you end up with disaster. This is the major cause of mistakes and problems that show up do to lack of long term testing.

More often than not I'm given a problem, I come up with a way to solve it, and I am told that the company won't pay for it and I need a cheaper way of doing it. I tell them that they can do it right or they can do it cheap and pay for the consequences later. They typically only accept the cheap band-aid fix.

[tye1138]

I

Just to start things off (oh, I'm sure I will piss off all of the Honda guys) I'll throw out the horrible valvetrain design of the '02-'08 CRF450R/X. The first time I worked on one of these was in '02. It wouldn't start (a soon to be familiar theme with these bikes). When I poped the valve cover off I was mortified that the entire valve train was held to the head with four 6mm bolts that were very long (hense 'very strechy'). The extra long 6mm bolts are more like 6mm springs, allowing the cam assembly bounce and hop on the cylinder head. The other shocking observation was the suprisingly small (narrow) cam bearings used. After doing a few of these top ends I discovered that the ones that were the chronic valve eaters all had one thing in common; the can bearings were worn enough that I could get up to .003" radial movement in a ball bearing. It was usually the cam sprocket side, the one that is not replacable. The effect is obvious; you can't posibly set the valve clearance correctly with worn out cam bearings.

Honda does some strange stuff because they're all about prototyping. They take a unique design and force it on the market to show their engineering prowess. Most of the other brands focus on copying each other and also stick to engineering that is more based in common sense. I mean cam's should be held in place by solid bearings with high pressure oil lines lubricating them. Ball bearing's will never be able to hold cam's properly because of the flex characteristics of the cam itself. I know quite a bit about Ducati engines and their first modern 4 valve motor which was made in the mid 80's called the Desmoquattro had ball bearing cam holders. The motor's revved to between 9500 - 13,000 RPM (depending on if street or race cam profile) and at 9500 RPM, the cam's bent so much, the cam would hit the rocker arm on the side instead of square and literally scrape off the chrome coating, imagine at 13,000 what damage was done. Ducati spent years trying to figure out and resolve the problem, which was pretty straight forward. When they did figure it out, they changed the head to use solid bearings and the problem pretty much went away with the newer testastretta head.

I've worked on a lot of motors from japanese sport bikes and Ducati's to Japanese 2/4 stroke and KTM dirt bikes. I've seen a great deal of engineering gone wrong and it really upsets me. What upsets me the most is how engineers don't ever think about the user doing service. They don't necessarily build frames so its easy to get the head off without taking the motor out. They use archaic and very simple designs to do complex tasks like chain driven cam systems, which I think is stupid in todays age. I've seen more motor failures from cam chain tensioner failures, then almost anything else! So guys, get rid of the cam chain!!! Make it gear driven, this is NOT rocket science! ?

I think what manufacturers should do build a prototype and then have someone live with it for a while. Ya know, do a valve job, pull the motor out, etc. Then give feedback to the engineer's, show them what would make it better and then, fix it! I wish I had the money to start a US built motorbike company; pavement and dirt machines. I could make such amazing little motors and chassis that make sense. Yep, they'd be slightly more money then the junk we already have today. But you spend that money for quality, stuff that doesn't blow up!

End rant...

[adam728]

Simple things like fasteners are never even brought up, so most new engineers never bother to think about how things are actually going to be held together.

We spent a lot of time on nuts/bolts, bearing, welded connections, springs, etc at Michigan Tech. Still reference one of my 10 year old book probably once a week.

[2grimjim]

I think the design flaws were a result of Honda seeing a marketing opportunity in the complaints against the YZ400/426's weight. Despite the fact that that the Yamaha was considered light by 4 stroke standards, there were endless articles in MXA, Dirt Rider, and Dirt Bike on how to shave weight from them (and spend buckets of money). Remember, in '02 when the Honda CRF came out, it was nearly 20 lbs lighter than the 426 and, at the time, probably the lightest production 450 4 stroke to ever be produced.

I think marketing forces more than anything else drove the engineering in that case.

As far as building motorcycles in the USA, I couldn't agree more. Not that I'm unhappy with the current offerings, but it's a shame that the way we have let manufacturing work be left to the chinese. I was very hopeful when Canondale stepped up to the plate and decided to enter the motorcycle market and very disappointed when they failed. I think the fatal flaws learned from that could serve as a valuable lesson to someone that wishes to build motorcycles in the USA. Canondale tried too hard to create something 'different' (even though they were way ahead of their time, look at the new YZ), and for whatever reason found more conventional designs less than honorable. Look at some of the tiny dirt bike manufacturers in Europe; Husaberg was started by a handfull of disgruntled ex-Husqvarna employees that were fired when Husqvarna was bought out by Cagiva; TM, Vertamati, VOR, and a few others are very small operations that made fairly conventional machines and only sold a handfull of machines every year yet they are still in business. Husaberg seldom sold more than 1500 machines a year worldwide before their purchase by KTM.

I think if you were to model a manufacturing operation similar to one of these small manufacturers it could be successful. Canondales mistake was to have a vision of competiting with the 'big boys' before the first engine made a spark. You can't convince me that someone in Japan had their attention focused on what Canondale was doing and there was money that changed hands behind closed doors to see that there was bad press coverage of the bikes that wasn't deserved. I really don't understand what ATK is doing at the moment. They are wheeling and dealing, and trying to import Hyosung (formerly imported as United Motors, a.k.a. UM into the USA). I just can't figure out what the White brothers are up to.

I think a small manufacturer could succeed by flying 'under the radar' and just produce Americam made machines for the US market. All you need to do is produce enough machines to comply with homologation rules and you could have a fighting chance to succeed.

[36MotoMarc]

You can't convince me that someone in Japan had their attention focused on what Canondale was doing and there was money that changed hands behind closed doors to see that there was bad press coverage of the bikes that wasn't deserved. .

.

Wait a minute. You use Honda's Unicam, an out-of-the-box yet successfull design, as an example of poor engineering, but them try to prop up Cannondales POS design (on SOOOO many levels) as great but only failed because the MAGS were bought out????!!!!!!!! Pfffff.

[ickfinger]

How about the first kx250f's that you had to drain the coolant to change the oil filter due to the one piece filter/h20 pump cover?!

[2grimjim]

Wait a minute. You use Honda's Unicam, an out-of-the-box yet successfull design, as an example of poor engineering, but them try to prop up Cannondales POS design (on SOOOO many levels) as great but only failed because the MAGS were bought out????!!!!!!!! Pfffff.

Hmmmm.....like I said, I'll probably piss off some Honda owner.

Lets look at what Canondale was doing (and made work) in 2002;

The Canondale had fuel injection, that worked.

It had a layout with a reversed cylinder head for better mass centralization (ummm... I don't think I need to explain that).

Had top top notch suspension components, albeit not set-up for motocross purists.

Electric start on a bike intended for pure competition (and worked better than some others like Husaberg and KTM electric 'restarters').

Only the 2nd maufacturer to use an Aluminum frame (OK, it was a copy of Hondas).

I'm no aplolgist for Canondale. They did sell an unfinished product to the public amid self-inflicted public hype. That was there own fault by making big promises to investors that quickly became unhappy when they weren't turning a profit in 18 months.

I test rode a Canondale in 2003 and was impressed with it. I'm aware of the crank issues they had, the programming bugs in the ECU, and some of the other quality issues.

The failures of the Canondale were bad business decisions. All of the hardware problem were eventually ironed out with the machine but the damage was done.

The bad publicity in the magazines was also the fault of Canondales marketing geniuses. Why on earth would you send out a machine that was unfinished to be evaluated? That was obviously from pressure to satisfy investors. I'm not trying to make any claim that the Canondale was without flaws, it was simply under developed. The problems with the magazines is that they gave Canondale no quarter in the reviews. Rather than just quietly hand the bike after testing back to Canondale an tell them that they had some issues and understanding that company was new to the business and trying to foster a positive relationship with the company and offer assistance, they chose to stand on their soap box and shout at the top of their lungs that the Canondale was a huge pile of s***.

Magazines profits come from advertizing. Follow the money. I put as much faith in magazine reviews as I do in used car salesmen, insurance adjusters, recruiters, and politicians.

The Uni-Cam is hardly 'out-of-the-box' engineering, and the way it was executed in the 250R was a huge improvement in the basic design originally used on the 450. But there are still a lot of pissed off ex-Honda riders that got real tired of putting a new set of valves in their bike twice a season. Granted you can make a 1st gen 450R live a long and happy life, they require extrordinarillary perseverant maintainence to do so. Something that is the direct result of flawed engineering.

[Hartebreak]

Z spokes

[2grimjim]

Z spokes

2 spokes in 1?

[2grimjim]

I mean cam's should be held in place by solid bearings with high pressure oil lines lubricating them. Ball bearing's will never be able to hold cam's properly because of the flex characteristics of the cam itself. I know quite a bit about Ducati engines and their first modern 4 valve motor which was made in the mid 80's called the Desmoquattro had ball bearing cam holders. The motor's revved to between 9500 - 13,000 RPM (depending on if street or race cam profile) and at 9500 RPM, the cam's bent so much, the cam would hit the rocker arm on the side instead of square and literally scrape off the chrome coating, imagine at 13,000 what damage was done. Ducati spent years trying to figure out and resolve the problem, which was pretty straight forward. When they did figure it out, they changed the head to use solid bearings and the problem pretty much went away with the newer testastretta head.

Not all ball bearings are created equal. Bearings are rated by class; class 1 through 7. A class 1 bearing is something you would find in a cheap axle bearing, a class 7 bearing in the spindle of a Bridgeport mill. Most bearings used in motorcycle engines are class 3 or 4.

[ikis86]

little story that always made me smile, about the notoriously unreliable transmissions on the Tiger tanks during WW2. They didn't find out until after the war that the forced labourers were machining groves into the main shafts to weaken them. could be classed as an engineering disaster i guess.

[36MotoMarc]

You're argument for Honda valvefailure being associated with the cambearing or bolt size for camtower has NO BASIS IN FACT. However Cannondale clearly missed the basics of BOTH exhaust length, and intake tract design, provable. To argue that one was bad engineering, while the other was only "bad marketing" or a payoff is absolutely absurd and insulting. It would be more credible to hack on the Cannondale.

"Cannondale's mini-headpipe exhaust" Check.

"Cannondale's choked-frame-intake-requiring-second-bandaid-airfilter intake" Check!

Honda's Unicam? Not mentionable in the same paragraph.

Several magazines dubbed it "Bike of the Year!!!!!!!!!!!". Didn't help it work any better, it was still a p o s. A pretty, glamorous, overhyped POS.

A PEFECT SUBJECT FOR THIS THREAD!!!!

The "Z" spoke was garbage!! Wasn't that Yamaha, circa '86?

The KXF (RMZ) Oil change debacle was a joke!

Edited January 5, 2012 by 36MotoMarc

[grayracer513]

The Cannondale was a poorly executed, half-assed attempt at being innovative for its own sake, and as an excuse for not doing things the way they should be done.

The reason the head was turned around had nothing to do with mass centralization. The cylinder is canted forward in spite of it. The fact is that when they insisted on using the twin beam spar frame design along with a forward canted shock without linkage, there was nowhere else to put the intake. Then they topped that off by forcing it to inhale around the steering stem. ? And yes, an electric start on a bike that was intended for competition...and weighed 280 pounds. Cannondale killed the Cannondale all by themselves, just by building it.

As to the Honda CRF450, I doubt there is any element in the Uni-Cam design that contributes in any way to the valve problems that engine has. They simply never used the right combination of valve material, valve coating, and seats. Incidentally, the Isuzu 4 cylinders used in the Geo line-up by Chevrolet had what amounts to Uni-Cam engines in them. The cam sat directly over the intakes an operated the exhaust with rockers. They were OK.

[tye1138]

Not all ball bearings are created equal. Bearings are rated by class; class 1 through 7. A class 1 bearing is something you would find in a cheap axle bearing, a class 7 bearing in the spindle of a Bridgeport mill. Most bearings used in motorcycle engines are class 3 or 4.

Of course, but it wasn't the ball bearings necessarily causing any issues, it was the fact the solid bearing held more of the cam at a tighter tolerance. I guess you could just put a big-ass ball bearing in the same place, but it wouldn't have the same effect. Its funny because I think we'll see more and more main crank bearings being solid bearings in the future as its more reliable and with better metallurgy/manufacturing processes, its going to be a huge step in making a more reliable motor.

[grayracer513]

The KXF (RMZ) Oil change debacle was a joke!

Yeah, who thought that one up?

[1987CR250R]

You have to sit down and ask yourself why about all of these engineering questions. Why combine the oil filter and water pump in one housing? Obviously weight and cost are top answers. But there is a lot more to the question. What is the design life of the engine? How often does Kawasaki recommend doing top ends? 20 hours? How often do they recommend changing the oil filter? 20 hours? Most bikes don't recommend changing the filter at every oil change. So if you're going through the head at 20 hours you're dumping the coolant at 20 hours, so that would be a good time to change your filter.

This wasn't a case of bad engineering it's just a case of engrained practices and customers who aren't willing to change their ways.

How about the ball bearings on Honda's Unicam? Use an ABEC-5 bearings? They're expensive, well over $100 for a bearing that size. ABEC-7? They're preloaded and come in matched pairs of angular contact bearings. That means they don't run free, you need two to do the job of a single deep groove bearing, and you're looking at well over well over $100. Journal bearings are cheap and have much longer life expectancies in similar applications. Why did Honda not use the journal bearing? Journal bearings require substantially more oil flow to stay lubricated. That requires a larger pump which is parasitic to horsepower. But also you have a pump pushing oil to the head under pressure but it has to return by gravity. With such a wide range of engine operating speeds seen on a MX engine you have to maintain enough oil flow at low rpms while not having too much at high. It's easier just to use ball bearings which require hardly any oil at all at any speed.

There are two sides to every coin. Try to think about what the engineers were thinking when they designed it and not just what they weren't thinking. And remember... it's all a game of cost. Saving a dollar on the head lets you spend a dollar on your forks and that may be what gives you the advantage over our competition while still meeting the same price point.

[Hans Schmid]

Any high performance 4 stroke (250f, 450f), that was/is a engineering disaster.

[jar944]

Its funny because I think we'll see more and more main crank bearings being solid bearings in the future as its more reliable and with better metallurgy/manufacturing processes, its going to be a huge step in making a more reliable motor.

That will never happen, plain bearings need a steady supply of pressurized oil. What you fail to realize is that unlike in a automotive application a dirt engine may end up out of position, (upside down) or any one of a long list of other ways to lose pressure.

A roller/ball can survive with minimal/splash lubrication, a plain bearing cannot.

People would complaining of the plain bearing engineering disaster and be angry about buying new cases and crank when you lost oil pressure, rather than new bearings.

[2grimjim]

Ahhh! I love stirring up a hornets nest! Before anyone has a stroke.......

How 'bout this one. Had a feller offer me a bunch of money to restore this bike and after I did a little research found out he could pick up one in good original condition for half the cost.

Does anyone know what this bike is and it's engineering significance?