Open Source Engines/Frames

byke
Posts
3033
Joined
8/12/2015
Location
Auburn, CA US
5/11/2018 1:14pm
It'd be funny if the bottom of that upper triple clamp were flat, because then it'd just be an upside down triple clamp.
Luxon MX
Posts
1381
Joined
11/6/2017
Location
San Diego, CA US
Fantasy
5/11/2018 1:32pm Edited Date/Time 5/11/2018 1:33pm
byke wrote:
It'd be funny if the bottom of that upper triple clamp were flat, because then it'd just be an upside down triple clamp.
Not quite flat. These are the Yamaha clamps, top and bottom of the upper and lower clamps:



The topology optimization results in a very non-intuitive design. These ended up coming out with essentially the same stiffness, but 25% stronger and 0.8 lb lighter than the stock clamps.
5/11/2018 1:39pm
Already Been Done......Team Hurricane baby....even Rick Melon made his privateer version




byke
Posts
3033
Joined
8/12/2015
Location
Auburn, CA US
5/11/2018 1:41pm
Bottom two are the upper clamp, bottom on the right and top on the left?

Two above those are the bottom clamp, bottom on the right and top on the left?

Insanely trick looking for triple clamps.

The Shop

Pigspit
Posts
41
Joined
7/16/2017
Location
Post Falls, ID US
5/11/2018 1:45pm
byke wrote:
It'd be funny if the bottom of that upper triple clamp were flat, because then it'd just be an upside down triple clamp.
Luxon MX wrote:
Not quite flat. These are the Yamaha clamps, top and bottom of the upper and lower clamps: [img]https://p.vitalmx.com/photos/forums/2018/05/11/260778/s1200_20180511_132623.jpg[/img] The topology optimization results in a very non-intuitive...
Not quite flat. These are the Yamaha clamps, top and bottom of the upper and lower clamps:



The topology optimization results in a very non-intuitive design. These ended up coming out with essentially the same stiffness, but 25% stronger and 0.8 lb lighter than the stock clamps.
Would love to see the stress analysis on those if you are willing to share. Beautiful parts. I am using Inventor Fusion, so I don't have access to the analysis stuff that 360 can do. Did some FEA with Pro/E back in engineering school. Maybe it's time to upgrade.

Homey55
Posts
1259
Joined
2/18/2010
Location
collinsville, OK US
5/11/2018 1:48pm
.kyle wrote:
Robot Bike are 3d printing titanium frame lugs for their mountain bikes. There is also some other company 3d printing dropouts for their bikes. I believe...
Robot Bike are 3d printing titanium frame lugs for their mountain bikes. There is also some other company 3d printing dropouts for their bikes. I believe both use the Renishaw machines.

Not sure how I'd feel just yet about 3d printed metal mx parts, esp in critical areas.
GM is printing seat brackets for their vehicles...
Luxon MX
Posts
1381
Joined
11/6/2017
Location
San Diego, CA US
Fantasy
5/11/2018 2:06pm
byke wrote:
Bottom two are the upper clamp, bottom on the right and top on the left? Two above those are the bottom clamp, bottom on the right...
Bottom two are the upper clamp, bottom on the right and top on the left?

Two above those are the bottom clamp, bottom on the right and top on the left?

Insanely trick looking for triple clamps.
Correct, and thanks!
philG
Posts
10958
Joined
5/12/2012
Location
GB
5/11/2018 2:28pm
.kyle wrote:
Robot Bike are 3d printing titanium frame lugs for their mountain bikes. There is also some other company 3d printing dropouts for their bikes. I believe...
Robot Bike are 3d printing titanium frame lugs for their mountain bikes. There is also some other company 3d printing dropouts for their bikes. I believe both use the Renishaw machines.

Not sure how I'd feel just yet about 3d printed metal mx parts, esp in critical areas.
Homey55 wrote:
GM is printing seat brackets for their vehicles... [img]https://p.vitalmx.com/photos/forums/2018/05/11/260780/s1200_gm_next_gen_lightweighting_proof_of_concept_part.jpg[/img]
GM is printing seat brackets for their vehicles...
Renishaw are doing some cool stuff, as are Matsuura, who have 3d Printing and % Axis machining in the same machine, which lets you do stuff you just cant print,

Renishaw did this bike a while back , was quite heavy though .

Johnny Depp
Posts
6435
Joined
10/16/2014
Location
Buda, TX US
5/11/2018 2:49pm Edited Date/Time 5/11/2018 2:50pm
There is a karting company out of Canada making CR500 engines, with some upgrades. BRP?
Rushton597 wrote:
Pigspit wrote:
Good find. Those motors are ~$6,000, fully assembled.
CR500 engines have been made into all sorts of radical stuff, it would mostly never be cost effective for MX, hence building custom cases like Service Honda did made the bikes too damn high priced. Cool stuff to look at though:

twin

V-four



GTCD15317
Posts
105
Joined
9/4/2017
Location
Burgettstown, PA US
5/11/2018 3:16pm
TTR's are made in china...... there ya go
a22
Posts
873
Joined
7/21/2011
Location
London GB
Fantasy
5/11/2018 3:29pm
.kyle wrote:
There was a person on Vital a couple of years ago who made his own mx bike and posted a topic about it. I think he...
There was a person on Vital a couple of years ago who made his own mx bike and posted a topic about it. I think he came from Europe. I just wish I could remember the name of the bike, but it looked like a good effort!
JM485
Posts
5800
Joined
10/1/2013
Location
Davis, CA US
5/11/2018 4:41pm Edited Date/Time 5/11/2018 4:46pm
Since this morphed into a but of a 3d printing topic, I figured I would throw this out there since it may not be obvious to people who are not familiar with 3d printing. One of the real world benefits to additive manufacturing (which is what 3d printing is), is that shapes you could not physically make from subtractive manufacturing (such as machining, etc.) are highly possible.

Consider what is know as the "infill". In most prints, it's not necessary or beneficial to print the part as a 100% solid. It's generally just a waste of material, and printing from about 75% infill to 100% infill yields an increase in part strength that is all but negligible. Now, to explain a little bit further, since we don't want prints to be completely hollow, the inside of a 3d printed part is filled with what is aptly named "infill". Now, there are multiple patterns widely used, but essentially it boils down to some sort of grid like pattern printed within the part to provide support between the outer faces, or "shells". The implications of this are just mind blowing, as it's possible to create parts that are significantly lighter than machined pieces, but still hold a significant amount of strength. This also reduces material used in the part, and eliminates the wasted material inherent in subtractive manufacturing.

This image gives a great example of different infill percentages, and you can probably figure out intuitively why it doesn't make much sense to go over 50% infill in most applications.

Other possibilities are that supports can be changed anywhere within the part, such as more support in a higher stressed area, and less support in a low stress area. This allows for the most optimized part as far as weight and function are concerned.

Before parts are printed, the solid model is taken from the CAD program and run through a slicing software. This does exactly like it sounds, cuts the model into many small slices, the thickness of which is set by the user. We usually print at .2mm layer heights, but they can go down to much higher resolutions, it really comes down to a balance between print quality and speed. The slicing software is where you can edit all of the infill settings and shell thicknesses, as well as tons of other settings I'm not experienced enough to know what to do with yet.

It's a really cool technology, I've already used our work printer to make prototype moto parts to test fit and it works awesome! These will absolutely be a necessity in any engineering office or department within the next ten years, now that we have one I don't know how we lived without it!
CG118
Posts
650
Joined
6/27/2014
Location
Kennesaw, GA US
5/11/2018 5:56pm
.kyle wrote:
There was a person on Vital a couple of years ago who made his own mx bike and posted a topic about it. I think he...
There was a person on Vital a couple of years ago who made his own mx bike and posted a topic about it. I think he came from Europe. I just wish I could remember the name of the bike, but it looked like a good effort!
EZZA 95B wrote:
Hans? [img]https://p.vitalmx.com/photos/forums/2018/05/11/260681/s1200_CR_Two_Two_Prototype.jpg[/img]
Hans?
I laughed out loud when I saw this. Some beer went up into my sinus cavity.
Luxon MX
Posts
1381
Joined
11/6/2017
Location
San Diego, CA US
Fantasy
5/11/2018 6:29pm
Pigspit wrote:
Would love to see the stress analysis on those if you are willing to share. Beautiful parts. I am using Inventor Fusion, so I don't have...
Would love to see the stress analysis on those if you are willing to share. Beautiful parts. I am using Inventor Fusion, so I don't have access to the analysis stuff that 360 can do. Did some FEA with Pro/E back in engineering school. Maybe it's time to upgrade.

Sure, here are a couple screenshots. We use a program called Altair HyperWorks for all our optimization and analysis. It's dramatically more powerful than Fusion 360 and is pretty much the industry (engineering industry) standard for optimization.

This model is highly complex in order to capture the physics as accurately as possible. Full non-linear contacts, press-fit of the steering stem, pre-tension of the bolts, non-linear loadcase continuation, etc. all with multiple loadcases to capture every event the clamps can be subjected to. Additionally we model everything from the axle to the handlebars in order to avoid over-constraining the model. The first shot is the topology optimization result, which shows the optimum material layout in the part. From there we create a CAD model (we use SolidWorks) that is manufacturable, then re-analyze, and make minor adjustments based on the results of that analysis (second image).




MR. X
Posts
6913
Joined
6/24/2010
Location
North Tonawanda, NY US
5/11/2018 6:34pm
CR500 engines have been made into all sorts of radical stuff, it would mostly never be cost effective for MX, hence building custom cases like Service...
CR500 engines have been made into all sorts of radical stuff, it would mostly never be cost effective for MX, hence building custom cases like Service Honda did made the bikes too damn high priced. Cool stuff to look at though:

twin

V-four



The top V twin cases were machined in the guys garage . I posted that picture a few years back and someone was able to post a link to a website with a full build story on it.
5/11/2018 8:45pm
byke wrote:
It'd be funny if the bottom of that upper triple clamp were flat, because then it'd just be an upside down triple clamp.
Luxon MX wrote:
Not quite flat. These are the Yamaha clamps, top and bottom of the upper and lower clamps: [img]https://p.vitalmx.com/photos/forums/2018/05/11/260778/s1200_20180511_132623.jpg[/img] The topology optimization results in a very non-intuitive...
Not quite flat. These are the Yamaha clamps, top and bottom of the upper and lower clamps:



The topology optimization results in a very non-intuitive design. These ended up coming out with essentially the same stiffness, but 25% stronger and 0.8 lb lighter than the stock clamps.
Some nice engineering and machine work there. Crazy what machine shops can turn out these days. What predicted direction are the stress loads based on? I've seen a few of the late gen KTM clamps that are machined out to a minimum that are twisted from the load placed on them in a crash by the ultra strong handle bars that are around these days and they twist the clamps past yield and I think that is due to the lack of material and strength in them for that type of force. I think it's all based on up and down motion and forward and rearward forces applied to the wheel with some turning strength as a a pair, not against each other, i.e handlebar force one way and wheel force the other when they both hit the ground simultaneously in an end over end. Hard not to eliminate carnage in that scenario I know.
I know you have way more knowledge on this but this issue also has to be considered because nobody wants to lay down $300 for a nice set of clamps to lose 50g only to have them bent up in their first crash. Word spreads quick these days and being stuck with couple hundred grand of brightly colored butter is no way to retire like a fat cat.
Pigspit
Posts
41
Joined
7/16/2017
Location
Post Falls, ID US
5/11/2018 9:35pm
Luxon MX wrote:
Sure, here are a couple screenshots. We use a program called Altair HyperWorks for all our optimization and analysis. It's dramatically more powerful than Fusion 360...
Sure, here are a couple screenshots. We use a program called Altair HyperWorks for all our optimization and analysis. It's dramatically more powerful than Fusion 360 and is pretty much the industry (engineering industry) standard for optimization.

This model is highly complex in order to capture the physics as accurately as possible. Full non-linear contacts, press-fit of the steering stem, pre-tension of the bolts, non-linear loadcase continuation, etc. all with multiple loadcases to capture every event the clamps can be subjected to. Additionally we model everything from the axle to the handlebars in order to avoid over-constraining the model. The first shot is the topology optimization result, which shows the optimum material layout in the part. From there we create a CAD model (we use SolidWorks) that is manufacturable, then re-analyze, and make minor adjustments based on the results of that analysis (second image).




Thank you, absolutely incredible. I had considered asking you if you had to model the forks as well to do the analysis, but it didn't even occur to me that you could go as detailed as bolt tension and stem press fit. Truly impressed.
cwtoyota
Posts
2397
Joined
3/11/2013
Location
Tacoma, WA US
5/11/2018 10:10pm Edited Date/Time 5/11/2018 10:12pm
Luxon MX wrote:
Sure, here are a couple screenshots. We use a program called Altair HyperWorks for all our optimization and analysis. It's dramatically more powerful than Fusion 360...
Sure, here are a couple screenshots. We use a program called Altair HyperWorks for all our optimization and analysis. It's dramatically more powerful than Fusion 360 and is pretty much the industry (engineering industry) standard for optimization.

This model is highly complex in order to capture the physics as accurately as possible. Full non-linear contacts, press-fit of the steering stem, pre-tension of the bolts, non-linear loadcase continuation, etc. all with multiple loadcases to capture every event the clamps can be subjected to. Additionally we model everything from the axle to the handlebars in order to avoid over-constraining the model. The first shot is the topology optimization result, which shows the optimum material layout in the part. From there we create a CAD model (we use SolidWorks) that is manufacturable, then re-analyze, and make minor adjustments based on the results of that analysis (second image).




Pigspit wrote:
Thank you, absolutely incredible. I had considered asking you if you had to model the forks as well to do the analysis, but it didn't even...
Thank you, absolutely incredible. I had considered asking you if you had to model the forks as well to do the analysis, but it didn't even occur to me that you could go as detailed as bolt tension and stem press fit. Truly impressed.
This is not an "I told you so" comment.

This is the skill level I was referring to with regard to 3D modeling and design combined with metrology and manufacturing to back it up. It's one thing to draw up a part that looks good. It's an entirely different level to develop each component to match or exceed what's available from the factory. Ask Luxon how many hours he has in each of his clamps from concept to first article.

Just for starters, a basic seat of SolidWorks is priced in the same ballpark as a new motocross bike. The FEA stuff comes at an additional price and has its own learning curve.

Luxon, thanks for posting the screen grabs from your work.
How do you pronounce your company name? Is it Luck-Sawn or Loo-zawn ?
FiZ
Posts
8
Joined
4/12/2018
Location
AT
5/11/2018 11:58pm
A friend of mine Reverse engineered and optimized a Brake caliper from ap racing our formula Student Race car.
It's "printed" Titanium, but the machining afterwards was pretty difficult.

philG
Posts
10958
Joined
5/12/2012
Location
GB
5/12/2018 2:53am
FiZ wrote:
A friend of mine Reverse engineered and optimized a Brake caliper from ap racing our formula Student Race car. It's "printed" Titanium, but the machining afterwards...
A friend of mine Reverse engineered and optimized a Brake caliper from ap racing our formula Student Race car.
It's "printed" Titanium, but the machining afterwards was pretty difficult.

This is where the 'just because you can, doesnt mean you should' comes into play.. How long did that take to print and what did it cost to get it X Rayed?
Pigspit
Posts
41
Joined
7/16/2017
Location
Post Falls, ID US
5/12/2018 3:58am Edited Date/Time 5/12/2018 4:20am
Luxon MX wrote:
Sure, here are a couple screenshots. We use a program called Altair HyperWorks for all our optimization and analysis. It's dramatically more powerful than Fusion 360...
Sure, here are a couple screenshots. We use a program called Altair HyperWorks for all our optimization and analysis. It's dramatically more powerful than Fusion 360 and is pretty much the industry (engineering industry) standard for optimization.

This model is highly complex in order to capture the physics as accurately as possible. Full non-linear contacts, press-fit of the steering stem, pre-tension of the bolts, non-linear loadcase continuation, etc. all with multiple loadcases to capture every event the clamps can be subjected to. Additionally we model everything from the axle to the handlebars in order to avoid over-constraining the model. The first shot is the topology optimization result, which shows the optimum material layout in the part. From there we create a CAD model (we use SolidWorks) that is manufacturable, then re-analyze, and make minor adjustments based on the results of that analysis (second image).




Pigspit wrote:
Thank you, absolutely incredible. I had considered asking you if you had to model the forks as well to do the analysis, but it didn't even...
Thank you, absolutely incredible. I had considered asking you if you had to model the forks as well to do the analysis, but it didn't even occur to me that you could go as detailed as bolt tension and stem press fit. Truly impressed.
cwtoyota wrote:
This is not an "I told you so" comment. This is the skill level I was referring to with regard to 3D modeling and design combined...
This is not an "I told you so" comment.

This is the skill level I was referring to with regard to 3D modeling and design combined with metrology and manufacturing to back it up. It's one thing to draw up a part that looks good. It's an entirely different level to develop each component to match or exceed what's available from the factory. Ask Luxon how many hours he has in each of his clamps from concept to first article.

Just for starters, a basic seat of SolidWorks is priced in the same ballpark as a new motocross bike. The FEA stuff comes at an additional price and has its own learning curve.

Luxon, thanks for posting the screen grabs from your work.
How do you pronounce your company name? Is it Luck-Sawn or Loo-zawn ?
If you don't think the open source model can handle this level of engineering, you need to revise what you think open source means. This list includes a research satellite and several cars.

https://en.m.wikipedia.org/wiki/List_of_open-source_hardware_projects

In my profession, we build an maintain open source software because it is the BEST way to get the best eyes on the project. It isn't a DIY thing. Ever heard of Android? Linux? Firefox? Apache?

I am not surprised at all at the amount of work required to build things like this, because I do it. I am impressed with the quality of his work on a personal level.

It would take me years to figure out how to cast an engine case, but I could do the ecu portion of the project with a snap of my fingers. (not literally, but nearly that easy.)
Dtat720
Posts
1588
Joined
2/20/2015
Location
Flowood, MS US
5/12/2018 7:06am
Im all for open source projects. I own a cnc shop, have 17 mazak lathes and 4 Hurco mills and several hand mills and lathes, centerless grinder, etc. 23,000 sq ft of playground. I reverse engineer stuff all the time just to gain a better understanding and see if i can improve upon it. Thats how i came up with my bicycle hubs i have patented now.
Luxon MX
Posts
1381
Joined
11/6/2017
Location
San Diego, CA US
Fantasy
5/12/2018 9:27am
Pigspit wrote:
Thank you, absolutely incredible. I had considered asking you if you had to model the forks as well to do the analysis, but it didn't even...
Thank you, absolutely incredible. I had considered asking you if you had to model the forks as well to do the analysis, but it didn't even occur to me that you could go as detailed as bolt tension and stem press fit. Truly impressed.
cwtoyota wrote:
This is not an "I told you so" comment. This is the skill level I was referring to with regard to 3D modeling and design combined...
This is not an "I told you so" comment.

This is the skill level I was referring to with regard to 3D modeling and design combined with metrology and manufacturing to back it up. It's one thing to draw up a part that looks good. It's an entirely different level to develop each component to match or exceed what's available from the factory. Ask Luxon how many hours he has in each of his clamps from concept to first article.

Just for starters, a basic seat of SolidWorks is priced in the same ballpark as a new motocross bike. The FEA stuff comes at an additional price and has its own learning curve.

Luxon, thanks for posting the screen grabs from your work.
How do you pronounce your company name? Is it Luck-Sawn or Loo-zawn ?
Pigspit wrote:
If you don't think the open source model can handle this level of engineering, you need to revise what you think open source means. This list...
If you don't think the open source model can handle this level of engineering, you need to revise what you think open source means. This list includes a research satellite and several cars.

https://en.m.wikipedia.org/wiki/List_of_open-source_hardware_projects

In my profession, we build an maintain open source software because it is the BEST way to get the best eyes on the project. It isn't a DIY thing. Ever heard of Android? Linux? Firefox? Apache?

I am not surprised at all at the amount of work required to build things like this, because I do it. I am impressed with the quality of his work on a personal level.

It would take me years to figure out how to cast an engine case, but I could do the ecu portion of the project with a snap of my fingers. (not literally, but nearly that easy.)
Thanks for the compliments!

It was a huge undertaking to get to this point. I've owned an engineering consulting firm for 11 years now, and back in 2013 I started playing with the idea of applying our engineering expertise to motocross. It's been a part time effort since then, but really ramping up in the last year or so. I did not keep an official time log, but there are 1000's of engineering hours into these parts now. Much of that overlaps into future parts, but all was necessary to make the first set of clamps (well, to this level at least).

And if I didn't own an engineering firm already, it would be impossible to start up all this just for motocross parts without already being independently wealthy or winning the lottery! It was mentioned that a seat of SolidWorks costs about as much as a new bike, which is true, but if you add up all the software and computer expenses for everything - CAD, CAM, analysis, computer upgrades, etc., I could be buying five or six new 450s annually! And that's before we start talking about manufacturing...

Regarding the open source idea, it was mentioned that there are lots of large scale open source projects. But when was the last time you saw someone driving around in an open source car that they made? I'm sure it happens, but it's certainly not mainstream. Linux, Firefox, Apache, etc. are all open source success stories, but the barrier to entry is just so much lower than that of a tangible product. You don't need crazy expensive software to develop new software and you don't need anything to "manufacture" the software either (aside from a lot of talented people working together). And most importantly, the end user doesn't need to do anything but install it and use it! And I don't mean to downplay the effort involved in open source software at all, as it's a huge undertaking on its own with 1000's of people working on it to make it happen.

But let's say I made my triple clamp CAD models publicly available along with all the supporting analysis, assumptions, design constraints, etc. Someone still has to make them! And at low quantities, sending these to a job shop would run you $2k or more per set of clamps. You could get a group together to buy higher quantities, but unless you bought A LOT of them, they'd be less expensive to buy from an aftermarket company from the start. Of course you could try and make them yourself, but realistically you would either need to be a very accomplished machinist with a machine available to you and a lot of free time, or the design would have to be dramatically simplified to the point where they're no better than stock parts. Now apply this to every part of the motorcycle and it becomes a near impossible task.

As software gets more and more available and as 3D printing becomes more mainstream with associated cost reduction, the idea of an open source bike becomes more of a reality, but in my opinion it's still many years away. I bet we might start seeing it trickle in though. An open source suspension link to fit KTMs (or something similar) isn't a drastic undertaking, though still a lot of work for the average Joe!
Luxon MX
Posts
1381
Joined
11/6/2017
Location
San Diego, CA US
Fantasy
5/12/2018 9:48am
Some nice engineering and machine work there. Crazy what machine shops can turn out these days. What predicted direction are the stress loads based on? I've...
Some nice engineering and machine work there. Crazy what machine shops can turn out these days. What predicted direction are the stress loads based on? I've seen a few of the late gen KTM clamps that are machined out to a minimum that are twisted from the load placed on them in a crash by the ultra strong handle bars that are around these days and they twist the clamps past yield and I think that is due to the lack of material and strength in them for that type of force. I think it's all based on up and down motion and forward and rearward forces applied to the wheel with some turning strength as a a pair, not against each other, i.e handlebar force one way and wheel force the other when they both hit the ground simultaneously in an end over end. Hard not to eliminate carnage in that scenario I know.
I know you have way more knowledge on this but this issue also has to be considered because nobody wants to lay down $300 for a nice set of clamps to lose 50g only to have them bent up in their first crash. Word spreads quick these days and being stuck with couple hundred grand of brightly colored butter is no way to retire like a fat cat.
Thanks! We do all the machining in-house as well.

Our models represent all the "normal" loads you'd expect to see. The handlebars are constrained and loads are fed into the wheel in all directions, which represent "normal" riding loads. Similarly we do this same process but constraining the steering stop to represent a crash.

The loading in our analysis is based on a few different things:

We reverse engineered the stock parts as well as a few aftermarket parts to see what loading they could handle before failure in each direction of applied load, we equipped the bike with accelerometers and a few other sensors to datalog the actual loading seen while riding, and we did various hand calculations to sanity check the results of the above. All in all we have a very good idea of the actual loads seen on the parts for the vast majority of scenarios the bike will experience (unfortunately I can't share those as it was A LOT of work to get them).

And we certainly don't want something breaking in a crash, but you have to make a compromise somewhere. If we designed these parts to never break, then they'd be absurdly heavy and they would just make something else break instead - It's a motocross bike, and it's going to crash. The goal of the compromise is to develop the best (lightweight, low flex, strong) parts that handle the vast majority of crashes without damage.

Stock KTM/Husky clamps are really light, but as you mention they are pretty flimsy and have issues bending, particularly the bar mounts. We designed our KTM/Husky clamps to be dramatically stronger than the stock clamps, much stiffer, and maintain the same weight. On the other hand, stock Yamaha (and everyone else really) clamps are the opposite - they're really strong, plenty stiff, but really heavy. Our Yamaha clamp design goals were very different from our KTM clamp design goals in that we aimed to maintain stiffness and strength, but reduce weight as much as possible.

Now the reality is that no matter what you make, someone is going to break it! All you can do is design the parts to handle the vast majority of crashes. But if someone crashes just right, then it is what it is. In that case, you preferably have a "mechanical fuse", which in our case is the bar mounts. This is tricky as you need to design the bar mounts to fail before the other, more expensive, parts, but not so weak that they fail/twist on tip-overs and minor crashes. We've gone through a few iterations of the lower bar mount on both sides - too weak and they fail from minor crashes, and too strong that in a major crash something else ends up breaking. This is all part of the long development process. With our current design we feel that we have a very good compromise on that.

Post a reply to: Open Source Engines/Frames

The Latest