Re: WillMill - HSM Benchtop Composite Mill
Hi William and other persons of interest - Had more of a think about the diagonal bearing arrangement.
1) When you place two bearings for instance on a cantilever you create an "analytical wrench' The wrench works about the centre of the bearings. When the bearings are on the front face of a gantry the centre of action is forward of the front face. This means the moment is eccentric to the gantry ie you have a global moment on the gantry and a local moment (wrench) on the gantry. The wrench contributes to lozenging which is the shear deflection.
2) The shear centre of a member is the place at which if you place a moment the member does not lozenge, the member twists but does not change local shape. The shear centre on a square section (or any closed section) is its geometric centre or centroid. So by placing the bearings on the diagonal the wrench centre is close to the shear centre meaning the member is most effective at resisting torsion. So maybe this can be improved further by ensuring the bearing centre is at the shear centre of the member if possible. Cheers Peter
Theory | C3.2 Shear Centre | Solid Mechanics II
https://www.civilengineeringterms.co.../shear-center/
Re: WillMill - HSM Benchtop Composite Mill
Hi Lookers - There is a slight hickup in figuring out the action centre of the square rail bearings. If the rails are the round type that can't support a moment then the bearing geometric center is close enough. If you are using square rails that support a moment then the bearing center is unlikely to be at the geometric center. The action center will be dependent on the structure stiffness and the cantilever length. If you are using FE then you can look at the bearing reactions and figure out its action line. If not then the geometric center is your best guess. Cheers Peter
Re: WillMill - HSM Benchtop Composite Mill
Here's some nice work - https://www.youtube.com/watch?v=q0RE-h1VDIg
My aim would be to achieve the same quality machine at 1/3rd the weight.. Peter
Re: WillMill - HSM Benchtop Composite Mill
Hi William - Enjoy sailing, I'm away for a week so won't hear much of me for a while. Peter
Re: WillMill - HSM Benchtop Composite Mill
Hi Peter - The simplified bearing I made is "bonded" to the linear rails. Good catch about the fact the they should able to slide in order to predict the deflection of the machine more accurately. I changed the contact type from bonding to sliding, and F360 gave me some weird and unexpected results. I'm sure I can figure out a workaround, but at the moment I feel like I have to get back to the machines design instead of going further down this rabbit hole. I believe that the current simplified bearing that I presented above will land me in the ballpark of realistic deflection of the linear bearings I intend to use.
#49 Regarding the deflection in the heave direction of the simplified bearing I designed, the deflection is the same as when it was compressed (load from top to bottom). I'm amazed by what you can achieve with some quick calculations Peter, I always find those very interesting. But with the calculated dummy material I'm not sure how your bearing would behave when the load comes from different directions? - the goal is to make it deflect somewhat equally in all direction (top, front, right). - On a side note I'd like to thank you for posting the document on bearing stiffness here on the forum.
I'm applying to become a mechanical engineer so that I too can make calculations like these and more!
#50 The ssvf machine frame is very interesting Peter. What a machine! I'm genuinely impressed by the ingenuity of what some engineers can achieve. - I didn't know that hydrostatic bearings exited for guideways, what an amazing piece of technology.
#51 Great explanation Peter, thank you for introducing me to the principle of shear center. This is very valuable information for someone like me.
Thank you it was a great tour on the boat. The sailboat was made of glass fiber, so I couldn't stop thinking about building boats with all my new knowledge on composites (no experience working with them yet) haha.
Cheers William
2 Attachment(s)
Re: WillMill - HSM Benchtop Composite Mill
Hi Will - The dummy bearing has to have the "cushion" material on all sides not just the top. The model was quick to check the hand calcs. The model should also incorporate a feature to make the car captive to the rail. I'd use the actual rail model and a conforming car to suit.
William when I graduated from Uni, FE was for the rocket scientists and us mech engs had to do everything by hand. You are in a period where all of these things are available on the desktop but you must understand the basics or the desktop software will lead you astray. You must do reality checks often... The first FE I did was a beam model of a wheelchair. It took me a week to calculate all the co-ordinates of the nodes and enter them into the software and it crunched several hrs to get an answer (Z80 8bit processor). Then once I got into serious FE work automeshers were very very expensive so I spent a few years building FE meshes manually. You learn a lot doing these things.
Now for some time I have been interested in plywood metal laminates. So I did a comparison of a 100x100x6mm thick aluminium extrusion vs a 100x100x16mm thick MDF box. Outside the box is 4mm of al or 2mm of steel. The corner rad of the extrusion is 12mm so the "plating" is in between the fillets ie its 76mm wide by 1200mm long. The MDF handles the shear flow within the section and the plating handles the bending. Both the Al and steel plated sections deflect the same as the 100x100x6mm al section. This type of structure is easy to do and can incorporate extra features and the weights are similar. So I think the next machine I build may go this way. Peter
The LHS image is the aluminium section... the pad at the front is the load patch. I have placed a 1000N (100kgf) load 200mm below this to torque the gantry. They both deflect 0.2mm