Bike Design

[TrialZonn]

Hello to everyone!

I'm writing this topic because I want to start designing some bikes/frames and I don't know where to start from.

I personally have much experience in sketching, I am taking drawing lessons the last 3 years and I can use some designing software such as Corel Draw (this is mostly for logos though and only for 2D objects but it has some very powerful tools) with ease. I have also all the equipment needed...

Usually I spend my time drawing scaled parts of buildings or sketching portraits etc but have no experience with bikes...

So...where should I start from?What do I need to know about bike geometry, materials, components etc to design bikes that could be fully ridable?

Is there any useful software for frames design (even on 3D maybe?)

Must I have some experience in mechanics too to make the bikes oparational?

Any other information about bikes/frames/components design would be really appreciated!

Please everyone that has experience give me some advice!Thanks!

[Shaun H]

Hello to everyone!

I'm writing this topic because I want to start designing some bikes/frames and I don't know where to start from.

I personally have much experience in sketching, I am taking drawing lessons the last 3 years and I can use some designing software such as Corel Draw (this is mostly for logos though and only for 2D objects but it has some very powerful tools) with ease. I have also all the equipment needed...

Being able to draw very well is a useful skill in pure aesthetical design however an understanding of the mechanics of the forces in trials and the stresses through components/frames/forks is going to be more important in designing a good product (although a product that looks very good but might be worse may actually sell well! Such is the fickle mindset of many markets).

Usually I spend my time drawing scaled parts of buildings or sketching portraits etc but have no experience with bikes...

So...where should I start from?What do I need to know about bike geometry, materials, components etc to design bikes that could be fully ridable?

You should be drawing inspiration from what's already out there, what exists in other markets (XC, DH etc) to come up with your own unique idea. Then you can begin research into materials and components to see they can work with/around your idea. Of course this being design and so open ended the process is sometimes completely reversed! But if I had no clue about trials that's how I'd go about it.

Is there any useful software for frames design (even on 3D maybe?)

I imagine you're aware of 3D CAD programs such as SolidWorks, AutoCAD etc. These are useful tools if you know how to use them. Some components are very difficult to model efficiently in CAD (i.e. it may be "easy" but your model can be too complex to allow proper analysis if you don't know how to model properly)

Must I have some experience in mechanics too to make the bikes oparational?

I would say ideally yes, you should be well versed in mechanics of objects to be able to create a market leading trials product. Such emphasis is on weight and strength now, a designer must know how to get the most out of their designs, materials and manufacturing methods.

[Stanners]

Im doing a similar thing at the mo

Id say start looking at other companies and look at their design/geo's as a bases and think of the advantages disadvantages of each part of the geo. BMX's now have higher bb's cause its easier to spin or chainstays are short so its easier to manual bunnyhop. You dont reeaallly have to draw it 3D but it is very useful to look at your final design without paying a fortune for a proto. Im just drawing the design in 2D autoCAD to get to grip of how its going to look and how im going to knock it up. Have a look at Rhino 3D if you dont have any access to AutoCAD which is good, if you have the time to get good on it http://www.rhino3d.com/. (I think theres a free download on it still?)

It is very helpful to know the mechanical side of things to know how the frame would be built in the end so you dont have some silly un-makable frame. Most people on here will be able to rattle off the universal sizes on frames like a 1.1/8 inch headtube and 135mm spacing of the dropouts on a standard mountain bike, but there are lots of geometrical decisions which have to be made so the frame is good to ride like lengths of the top and bottom tube and length from centre of the bb to the centre of the head-tube. You have to know what material to use to start with to know what thickness tubing to use...steel is smaller as it can take more pain....and the advantages of the properties of the materials like aluminum fatigues more easily than steel which is more forgiving and ti dents easily (i thk).........

Basically if you want to do this properly or want to make a career out of it do engineering prod design at college then go to uni and do materials science/mechanical engineering/prod design/sport engineering(or whatever its called.....

Or if you want to do it as a hobby - us this forum - use google - use wikipedia - ask people for help and opinions - i think if your brainy enough you should be able to knock out a half decent frame with just knowing the basics of what to do.

Hope this helps - probably didn't - Good luck

Rob

Edited November 27, 2009 by Stanners

[TrialZonn]

Many thanks to both of you I found what you said really helpful!

I'll have a try with AutoCAD and everything else and see how it goes.

About the university, I'm really into architecture that's why I draw mostly buildings etc...

Thanks again, maybe I'll post some of my work in the future here to be seen and criticized.

The overall process may have some good results on my drawing and mechanics experience.

[psycholist]

3D models are the only way to be sure everything will work properly (Without a hell of a lot of extra effort anyway). The main thing that has to be done with whatever design you arrive at is to make sure it's compatible with existing components. If you have a trials bike handy measure things like the actual wheel diameter and width (Chainstay clearance is a big design constraint in setting up trials frames), make sure there's crank clearance around Magura pistons. Setting the brake mounts to be centered on the rim is critical too.

Google have a 3D CAD package they allow free download here and it seems pretty good: http://sketchup.google.com/

I'm a fan of industrial packages such as ProEngineer (My favourite package - as used by Caterpillar, the Volkswagen Audi group, parts of Toyota, Santa Cruz Bikes, John Deere and SRAM), Solidworks (You can see this in action on Orange County Choppers), Catia, Microstation, Rhino3D, CADKey and Inventor (Made by the same company as AutoCAD but good for 3D) and am pretty sure AutoCAD is the work of the devil unless all you do are 2D drawings, in which case you're missing out on a lot of the benefits of CAD in the first place. If you get used to any parametric 3D package you can pick up others pretty quickly too - it's a pretty useful skill to have as well.

Edited November 27, 2009 by psycholist

[AdamR28]

pretty sure AutoCAD is the work of the devil unless all you do are 2D drawings, in which case you're missing out on a lot of the benefits of CAD in the first place. If you get used to any parametric 3D package you can pick up others pretty quickly too - it's a pretty useful skill to have as well.

Agreed 100%!

[TrialZonn]

I will try out one of these for sure and taking measurements from my trials bike is a nice idea too!Many thanks!

[Alex Dark]

I'm currently building a frame for my a level project completely from scratch....done loads of research and cad work, but cant really be bothered to type it all ou there

email me / msn me at alexdark2@hotmail.co.uk if you want a chat.

[Stanners]

I'm a fan of industrial packages such as ProEngineer (My favourite package - as used by Caterpillar, the Volkswagen Audi group, parts of Toyota, Santa Cruz Bikes, John Deere and SRAM), Solidworks (You can see this in action on Orange County Choppers), Catia, Microstation, Rhino3D, CADKey and Inventor (Made by the same company as AutoCAD but good for 3D) and am pretty sure AutoCAD is the work of the devil unless all you do are 2D drawings, in which case you're missing out on a lot of the benefits of CAD in the first place. If you get used to any parametric 3D package you can pick up others pretty quickly too - it's a pretty useful skill to have as well.

Edited November 27, 2009 by Stanners

[psycholist]

In a weird twist of events PTC, the company who own ProEngineer bought out the people who wrote ProDesktop (It probably wasn't called that till PTC took it over). PTC then started giving out free copies of ProDesktop for general use to hook people into 3D CAD just about when people's home PC's were powerful enough to run decent 3D models without shitting themselves. I taught ProDesktop for a couple of years. It's a great package for getting completely new people into CAD, but a complete pain in the hole to work in once your models get complex or assemblies get big (Strangely it works very quickly if you limit each subassembly to 5 or 6 parts and assemble subassemblies rather than putting everything into one assembly in a flat structure). It's also hideously bad at using geometric constraints in the 2D sketches used to make 3D sketches and it makes a lot of dangerous assumptions that cause regeneration failures in the models when you change dimensions, so basically as soon as a sketch doesn't do what you want it to it becomes a nightmare to fix.

ProEngineer at the time ProDesktop was being given out looked like old unix software - it sat in two windows, one with the model rendered in 3D and the other a column of buttons with text on them that changed as you selected options. This interface was actually pretty fast once you got to know it as most commands involved a set of clicks in a vertical line down the menu, but it didn't have the intuitive benefits of ProDesktop so it was very hard to learn.

The program itself was rock solid in terms of its reproduction and calculation of geometry. It's a solid modeller in CAD terms, so rather than making a shape that looks like what you're drawing, which is all computer gaymes need to do to keep things pretty, a CAD package calculates the volume and areas of the parts being drawn from first principles - this allows the programe to calculate for eaxmple: interference between volumes, centres of gravity, surface areas etc. as well as creating geometry that can be CNC machined, 3D printed or analysed using finite element analysis or computational fluid dynamics.

Storing of files was and still is excellent, it automatically stores each save as a new revision as well as storing a trail file, which holds each mouse click made in the program while it was running, allowing some impressive crash recovery as well as considerable automation of certain functions. It was even pretty stable running on windows 2000, XP and Fvista... Over the last few years ProEngineer has started to look more and more like ProDesktop (Though the unix style side menus still crop up when you go into more advanced feature creation) but become even more powerful in terms of the geometry support - for example the curves and shapes on Logitech mice are all designed in ProEngineer and CNC machined from the CAD design. I'm not sure whether a PC running ProE can model all the complex curves of a car say (The mathematical constraints to keep all the reflections from something like a car smooth and even are heavy going in ProE), but it can get pretty close. I'm not saying it's the best for all situations, but it manages to do everything I've tried with very little fuss - though the learning curve can be pretty steep at times.

The thing is I know this CAD package - all its competitors will have many of the same features or better optimisation for particular jobs. For something like designing a bike frame without too much analysis almost any CAD package at all will do.

[TrialZonn]

So....this is my first try to put some shapes together...

I tried to make a bashring (thought that this was the simplest thing to make on a trials bike) and here it is:

The most thing on it hasn't been measured it's just an idea.

I used Google Sketchup which I found easy to use but I'll borrow AutoCad 2008 from a friend and try it a little bit.

By the way...please tell me some opinions about the Bashring...

Many thanks for all your help.

[psycholist]

Unless that's for a very big chainring the hole in the centre isn't big enough for a threaded crank. It should be slightly over 1.37".

[monkeyseemonkeydo]

Unless that's for a very big chainring the hole in the centre isn't big enough for a threaded crank. It should be slightly over 1.37".

The most thing on it hasn't been measured it's just an idea.

[OD404]

The thing is I know this CAD package - all its competitors will have many of the same features or better optimisation for particular jobs. For something like designing a bike frame without too much analysis almost any CAD package at all will do.

This is the main thing, we're all comfortable with a certain package because it's what we're familiar with. I use ProE on a daily basis too, so that is the one I generally recommend.

Going a little off topic here...

However there is one thing about it (and every other CAD package out there as far as I'm aware) that is guaranteed to infuriate me, and that's it's in ability to draw a proper circle. By the very nature of the way it models curves, a circle is actually an ellipse which is fine for you designer types as the package knows you are drawing a circle and will always treat it as such, but for an analyst like myself it's a nightmare. This becomes an issue for me on a daily basis, as I spend most of my time doing thermal, structural and flow analysis of cylinder heads and when your valves are actually elliptical, modelling the heat transfer from them to the valve seat will always produce the wrong results. The solution to this is to delete the valves and seats and model them in the analysis software (Ansys) which is as difficult to model in 3d as AutoCAD!!

Rant over

[TrialZonn]

Unless that's for a very big chainring the hole in the centre isn't big enough for a threaded crank. It should be slightly over 1.37".

Yep I know...I will measure everything next time.That bashring was made with the only purpose to see how these kind of software works...Thanks for your replies by the way.

Oh I forgot to ask...Could anyone give me the dimension of the tube that holds the bottom bracket?What's the length of it?Is it different from frame to frame?

Edited November 30, 2009 by TrialZonn

[psycholist]

The BB has 1.37" diameter threads cut into it (The thread that screws into it will have a max diameter of 1.37") and will either be 68 or 73mm long. Unless you go for one of the new standards, where figures past 100mm can occasionally appear, but no trials bikes I'm aware of use anything except a 68 or 73mm BB shell.

[TrialZonn]

The BB has 1.37" diameter threads cut into it (The thread that screws into it will have a max diameter of 1.37") and will either be 68 or 73mm long. Unless you go for one of the new standards, where figures past 100mm can occasionally appear, but no trials bikes I'm aware of use anything except a 68 or 73mm BB shell.

Many thanks!