Home Made Bikes

Author Message
Posted: Jan 2, 2020 at 15:03 Quote
I agree, there are too many variables to do it for the open market, but for a one-off homemade bike which you will be riding, I would have thought it would be easier to integrate the calculation into the design. Afterall, the tyre is acting in series with the suspension, allbeit with very little damping but it does have a spring rate...

I'm guessing that the calculations are good for getting a starting point, but it really boils down to how the bike feels on the trails and making changes based off that!

Posted: Jan 2, 2020 at 17:03 Quote
jgmu97 wrote:
I agree, there are too many variables to do it for the open market, but for a one-off homemade bike which you will be riding, I would have thought it would be easier to integrate the calculation into the design. Afterall, the tyre is acting in series with the suspension, allbeit with very little damping but it does have a spring rate...

I'm guessing that the calculations are good for getting a starting point, but it really boils down to how the bike feels on the trails and making changes based off that!

You're overestimating the precision with which the chassis and dampers can be tuned, let alone the precision a home builder can realistically attain, especially without dozens of iterations.

Posted: Jan 3, 2020 at 1:27 Quote
R-M-R wrote:
jgmu97 wrote:
I agree, there are too many variables to do it for the open market, but for a one-off homemade bike which you will be riding, I would have thought it would be easier to integrate the calculation into the design. Afterall, the tyre is acting in series with the suspension, allbeit with very little damping but it does have a spring rate...

I'm guessing that the calculations are good for getting a starting point, but it really boils down to how the bike feels on the trails and making changes based off that!

You're overestimating the precision with which the chassis and dampers can be tuned, let alone the precision a home builder can realistically attain, especially without dozens of iterations.

That was why I also asked about how deeply builders analyse the structural performance of their frames. Yes FE analysis would give you an idea of frame stiffness and you could optimise stiffness with tube profiles etc, but there’s still going to be compliance in the system and the likelyhood is most frame builders won’t have access to that level of computer program anyway.

I’m guessing for dampers you’d need to get a custom tune where the damper could be run on a dyno which is again out of reach for most one-off builds?

Seems like the best approach is experimenting out on the trail with fork/damper settings and tyre pressures.

Posted: Jan 3, 2020 at 3:40 Quote
cedrico wrote:
how frame design and suspension setup should be adjusted for tire volume and pressure.
You are adjusting your tire/tire pressure to fit your frame and suspensions, not the other way around.
And after that your riding style will smooth all the technical details.

Posted: Jan 3, 2020 at 7:13 Quote
jgmu97 wrote:
That was why I also asked about how deeply builders analyse the structural performance of their frames. Yes FE analysis would give you an idea of frame stiffness and you could optimise stiffness with tube profiles etc, but there’s still going to be compliance in the system and the likelyhood is most frame builders won’t have access to that level of computer program anyway.

I’m guessing for dampers you’d need to get a custom tune where the damper could be run on a dyno which is again out of reach for most one-off builds?

Seems like the best approach is experimenting out on the trail with fork/damper settings and tyre pressures.

Standard practice for designers is to just do the surfacing of the bike, often with little to no FE, and let the factory handle the lay-up and wall thickness. A particularly low-effort design firm once told me "we just draw whatever looks good and the factory sorts it out with lay-up". This firm has done only road bikes for the past decade, where the double-diamond shape and rigid frames are nearly impossible to thoroughly screw up; when tasked with doing a mountain bike, their incompetence forced me to take over the detail work and dimensions of the design, which I normally don't do.

Although this represents the lower end of the spectrum of competence, it's not completely unusual and illustrates the extent to which the industry is just guessing. Marketers create an image of teams of top gun engineers doing countless FE simulations to perfectly tune their precision machines. This is marketing, not truth. As an example, look at the weights of frames within a category and you'll find frames with comparable materials that differ in weight by pounds, yet the failure rates are completely independent of weight. Bikes are nowhere near as refined or optimized as they're made out to be.

At best, a modest amount of FE is done, and many companies do none. Those that do, rarely have a specific target in mind; they're just looking to smooth out the stress profile by reducing the worst hot spots, not unify the flex profile and tune it to achieve some known, ideal value.

A bit of CFD may be done for high-end road bikes, but again, some of these accomplish about as much as an F1 aerodynamics team could do over a lunch break.

Posted: Jan 3, 2020 at 7:55 Quote
R-M-R wrote:
jgmu97 wrote:
That was why I also asked about how deeply builders analyse the structural performance of their frames. Yes FE analysis would give you an idea of frame stiffness and you could optimise stiffness with tube profiles etc, but there’s still going to be compliance in the system and the likelyhood is most frame builders won’t have access to that level of computer program anyway.

I’m guessing for dampers you’d need to get a custom tune where the damper could be run on a dyno which is again out of reach for most one-off builds?

Seems like the best approach is experimenting out on the trail with fork/damper settings and tyre pressures.

Standard practice for designers is to just do the surfacing of the bike, often with little to no FE, and let the factory handle the lay-up and wall thickness. A particularly low-effort design firm once told me "we just draw whatever looks good and the factory sorts it out with lay-up". This firm has done only road bikes for the past decade, where the double-diamond shape and rigid frames are nearly impossible to thoroughly screw up; when tasked with doing a mountain bike, their incompetence forced me to take over the detail work and dimensions of the design, which I normally don't do.

Although this represents the lower end of the spectrum of competence, it's not completely unusual and illustrates the extent to which the industry is just guessing. Marketers create an image of teams of top gun engineers doing countless FE simulations to perfectly tune their precision machines. This is marketing, not truth. As an example, look at the weights of frames within a category and you'll find frames with comparable materials that differ in weight by pounds, yet the failure rates are completely independent of weight. Bikes are nowhere near as refined or optimized as they're made out to be.

At best, a modest amount of FE is done, and many companies do none. Those that do, rarely have a specific target in mind; they're just looking to smooth out the stress profile by reducing the worst hot spots, not unify the flex profile and tune it to achieve some known, ideal value.

A bit of CFD may be done for high-end road bikes, but again, some of these accomplish about as much as an F1 aerodynamics team could do over a lunch break.

That's quite shocking then when you consider some of the marketing claims you read/see...

Thank you for some great industry insight!!

Posted: Jan 4, 2020 at 3:44 Quote
Build me a new front triangle Smile
new front build up

Posted: Jan 4, 2020 at 13:57 Quote
Meet me at the plaza downtown and let's hit the stair set. #freerideaintdead

But seriously, that's a high instant centre, though it does drop quickly. Direct drive or idler?

Posted: Jan 5, 2020 at 1:05 Quote
R-M-R wrote:
Meet me at the plaza downtown and let's hit the stair set. #freerideaintdead

But seriously, that's a high instant centre, though it does drop quickly. Direct drive or idler?
The linkage and rear triangle seem to be from a zumbi f44

Posted: Jan 5, 2020 at 1:40 Quote
ajax-ripper wrote:
The linkage and rear triangle seem to be from a zumbi f44

Good eye! So, no idler.

Posted: Jan 5, 2020 at 8:59 Quote
Yeah Zumbi f 44 on idler just a ss tentioner. Now its all done with the hope M4 on and a chain so just need some time to get out end test it Smile

Posted: Jan 6, 2020 at 4:35 Quote
Bigtiresrbetter wrote:
It's not a full frame yet, but I'm starting to get excited! This is the bottom bracket junction of a gravel bike frame that I've been working on for a while now. It's compression molded out of fiberglass for now, and I'll re-do it in carbon fiber hopefully in the near future. This part was made in 3D printed molds using a removable 3D printed core material.

Bottom bracket junction molded with removable mandrel

Bottom bracket junction molded with removable mandrel

Bottom bracket junction molded with removable mandrel

Bottom bracket junction molded with removable mandrel

Fiberglass plies for a 20 layer mandrel layup of the bottom bracket junction

Looks nice, i want to work with 3d printed moulds aswell, which material you use for it?

Posted: Jan 6, 2020 at 5:57 Quote
R-M-R wrote:
jgmu97 wrote:
That was why I also asked about how deeply builders analyse the structural performance of their frames. Yes FE analysis would give you an idea of frame stiffness and you could optimise stiffness with tube profiles etc, but there’s still going to be compliance in the system and the likelyhood is most frame builders won’t have access to that level of computer program anyway.

I’m guessing for dampers you’d need to get a custom tune where the damper could be run on a dyno which is again out of reach for most one-off builds?

Seems like the best approach is experimenting out on the trail with fork/damper settings and tyre pressures.

Standard practice for designers is to just do the surfacing of the bike, often with little to no FE, and let the factory handle the lay-up and wall thickness. A particularly low-effort design firm once told me "we just draw whatever looks good and the factory sorts it out with lay-up". This firm has done only road bikes for the past decade, where the double-diamond shape and rigid frames are nearly impossible to thoroughly screw up; when tasked with doing a mountain bike, their incompetence forced me to take over the detail work and dimensions of the design, which I normally don't do.

Although this represents the lower end of the spectrum of competence, it's not completely unusual and illustrates the extent to which the industry is just guessing. Marketers create an image of teams of top gun engineers doing countless FE simulations to perfectly tune their precision machines. This is marketing, not truth. As an example, look at the weights of frames within a category and you'll find frames with comparable materials that differ in weight by pounds, yet the failure rates are completely independent of weight. Bikes are nowhere near as refined or optimized as they're made out to be.

At best, a modest amount of FE is done, and many companies do none. Those that do, rarely have a specific target in mind; they're just looking to smooth out the stress profile by reducing the worst hot spots, not unify the flex profile and tune it to achieve some known, ideal value.

A bit of CFD may be done for high-end road bikes, but again, some of these accomplish about as much as an F1 aerodynamics team could do over a lunch break.

I once did a job for a firm making forks they wanted less weight in their carbon schedule 25 g of that was telling them not to use as much glue honestly some of these places make trash

Posted: Jan 7, 2020 at 3:42 Quote
@ThoBoe I still using PLA for the molds believe it or not. I've messed around with some of the composite filaments (PETCF AND PA6CF) but found them difficult to reliably print large parts with, and they're expensive. I'm using a room temp cure epoxy system so the low glass transition temp of PLA isn't really an issue.

Posted: Jan 7, 2020 at 11:02 Quote
Hey a question for any Canadian builders out there ,where are you getting your brazzing rod and flux from.

I can find some 3/32 rod at airspruce but i can't seem to find the appropriate flux.

Thanks in advance.


 
Copyright © 2000 - 2020. Pinkbike.com. All rights reserved.
dv56 0.010487
Mobile Version of Website