[GreenvilleDave]

Hi everyone:

This is my first post to the forums after lurking for some time, and finally deciding to start my own CNC machine build.

Like many people it seems, I decided to use 'skate' bearings for my leadscrew support bearings. Ideally, i'd like to use three per axis, two pre-loaded against each other on one end ( poor man's matched AC set ), and one on the other end.

This meant that I'd need to do a lot of hole cutting for 22mm ( odd here in the states ) bearings.

Over the past two weeks, I set about to make some bearing blocks that I can use to fix the bearings. My goals were:

• make lots of them quickly
• accomodate several mounting schemes
• cheap to make

After lots of deliberation, I decided to use RTV silicon to accomplish the "lots quickly and cheaply parts". I also wanted to experiment with the casting process.

I ended up making two different kinds of block-- one that has normal 1/4" holes, and another one that is a bit more compact, and has counterbored holes for socket head cap screws.

I have posted some pictures-- hopefully i've not messed that up

The process was:

I first made a master using a drill press and forstner bits, out of machinable wax. The RTV rubber holds amazing detail, so i polished the master and squared it up as best as I could. This took more than one try

Then, I made an RTV mold. This took more than one try also. Making a mold box is a pain, I ended up using Lego bricks to minimize the amount of mold rubber used.


Then i used urethane resin to make several ones. each one needs to be sanded on the back, since the mold is a 1-part mold.


I was fairly pleased with the results. The blocks are plastic, but they do not need to have high radial stiffness anyway. My idea is to 'sandwich' a piece of angle-aluminum between two of these to make a double bearing block, like shown.

I have ended this phase encouraged-- i guess my goal with this post is that someone else might find this technique useful. I was really pleased with the accuracy of the parts that i have ended up with!

[Khalid]

GreenvilleDave
Very Nice idea This is very good to those who have no CNC Machine... I made all of the MDF parts on my CNC Machine...

[GreenvilleDave]

Thanks!

Yes, blocks like these are very simple to make once you have a cnc, for sure.

Like you mentioned, I figured folks who dont have that capability might find the technique interesting.

Also had the vague notion that if some people use this approach, a set of molds would crop up by which people could share parts with each other to get that first CNC machine most people make up and running, so that they can make their 'main' machine.

[harryn]

Hi Dave - I really like your approach here. There was another member who posted something similar in concept, but I cannot remember who it was. For some reason, I think they might have even used the actual bearing and the screws to help design the mold rather than drill them out. - not sure.

In any event, it is a handy way to make bearing blocks - a task I am now facing for my cnc router project as well.

As you pointed out, making the master mold square is a bit tricky if you do not have machining tools already - I have just borrowed small drill press and a hand held circular saw, so the build is definitely in stages.

Can you provide a link or part number for the urethane you used ? Thanks

I don't suppose you would be interested in selling some of those urethane blocks - would you ? I can buy bearing blocks for about $ 15 / each, but it adds up fast.

Thanks

Harry

[GreenvilleDave]

Hi!

I bought the RTV and Urethane from smooth-on ( i am not affiliated with them):

http://www.smooth-on.com

For home use, lower viscosity works best, and you want a resin that mixes by volume instead of mass so you don't need a scale.

I used these two products:

rtv rubber: oomoo- 25

http://www.smooth-on.com/Silicone-Ru...136/index.html

for plastic, i used smooth-cast 300 resin, because the viscosity is so low:

http://www.smooth-on.com/Urethane-Pl...209/index.html

I also got some C-1508 ( same link as above ), since it has very low shrinkage, is very strong, and is very hard. I have not tried it yet, it is quite viscous when it pours so i'm not sure it will work in my molds.


I know what you mean about the blocks-- I went the same path, and like you, figured out that $15 x three per axis x 3 axis = too much.


Its just so frustrating for such a simple part to be so hard-- i'd love it if some other folks would make some molds of timing belt pulleys, couplers, and other $10/each parts that we could make for each to bootstrap a machine.

I'll see if i can take some pics of my installations and put together some documentation-- if so i'll post some for sale over on ebay. It wouldnt cost too much to throw a few together from my molds I wouldnt think. if it helps folks get started, I'm game.

[GreenvilleDave]

Hi, harryn:

I've listed a set of these on ebay, case you would like some

http://cgi.ebay.com/ws/eBayISAPI.dll...m=190284005505
Dave

[lovebugjunkie]

Dave

You mentioned C-1508 being very strong. Is it strong enough if to made ¼ inch thick joining plates for 8020.

Something like this: http://cgi.ebay.com/80-20-T-Slot-Alu...742.m153.l1262

I was thinking of making molds out of cast acrylic but not sure if the C-1508 had the strength or even if it would be cost effective. Some of 8020's bigger plates and “L” sections can add up in cost with the number and type I was considering.

george

[GreenvilleDave]

Hi junkie,

I think that it would depend on the loads you are putting on the plate.

C-1508 has a tensile strength of 6000 psi, and flex modulus of 86ksi,

The absolute strongest urethane material i see available would be TASK-9 resin, with 7.8ksi tensile ,11ksi compressive strength, and 350ksi flex. modulus

To give some sense, aluminum has a tensile strength around 40ksi, and youngs modulus ( E ) of 10,000 ksi.

Plastics do not have a linear stress strain relationship, so you have to consider the load situation to compare the materials.

In a bending application , aluminum is 10,000/86 = about 100 times stronger than c-1508, and about 10,000/350 = 30 times stronger than Task-9

In tensile/compression, aluminum is 40/6 = 6 times stronger than C-1508, and 4 times stronger than task-9

so, in short you could expect quite a bit of deflection in bending compared to aluminum.

Looking at the joining plate you showed for example, it looked to be about 1/4" thick and about 3" wide, cross-sectional area is 0.75 in^2

Consider a 10 lb load at the end of a 3" piece of 80/20 connected to that plate. If we model that as a cantilever beam, ( with the plate in bending ):

The applicable formula is

deflection = load * length^3 / E ( modulus of elasticity ) * I ( area moment of inertia).

the area moment of inertia of that section is bh^3/ 3 = 0.25^3/3*3 = 0.015625 in^4

for aluminum, the deflection would be about 0.0005, ( based on modeling the plate as a cantilever beam ). This is based on the assumption that the deflection is lower than the yield strength of aluminum and thus in the linear range ( which it is).

in the same case, a c-1508 piece would deflect about 0.066.

for bearing blocks, the load is compressive, and operating in the that load case. for a 'skate' bearing, the a radial load is spread across an area of anout 0.75"x0.25" = about 0.18 in^2. at 6000psi, and a 2x factor of safety, we could handle 6000/2*0.1875 = 500 pounds of radial load. thats plenty given that most all the loads on the leadscrew are axial anyway, and deflection will have a small impact on axis accuracy.

hth