New Machine Build- 5' x 9' Torsion Box CNC with 1/4" CRS Rails

[ivojnob]

After a long time lurking, I've (like many others) finally taken the plunge and acquired most of the parts for my first router build. My plan is to be able to cut full sheets of MDF with a little wiggle room on each side. Because of the cost of materials and shipping (to Canada) I sadly decided against 8020 and have opted instead for a torsion box out of 3/4" MDF. Here's a shot of what I'm planning for the main base.

[ivojnob]

I've already received a fairly complete package of parts from Ahren, including 5 extended carriages, 3 standard carriages, 3 NEMA 23 rack and pinion drives, a G540, parts for my Z-axis, etc, etc.

I'm planning to run my X-axis rails vertically, similar to the arrangement of angles that Joe uses on his Hybrid. This is going to mean some changes from the way the carriages are usually used, but I think it will work. The CRS that I bought for the X and Y are 6" tall (to fit the 5" thick torsion box). This should help stiffen the box over its 9' length as well as the 5' wide gantry. I'm running dual R&P drives on the X, and a 5 start 1/2" ACME screw for the Z. Nothing groundbreaking here, but I really wanted to make the machine big enough from day 1.

Here are some pics of the design. I just started 3 months of parental leave from my job today, so my access to SolidWorks has come to an end (thus some differences between the pictures that I saved at different times over the last month or so). I might try to pick up the design process in Sketchup, but I'm getting the wood for the torsion box in a few days, so I'll likely skip more modeling and start posting build pics.

Let the criticism begin!

[DHSYCNC]

pulse equivalency，working speed, spindle power, accuracy is very important. if you choose pinion and rack and want to do 3D carving, surface working piece is not very smooth. it's better for you to choose ball screw in x axis because x is main axis when engraving. Usually 3/4'' MDF can be cut off by 3-5kw spindle

[DHSYCNC]

you can use Artcam or other CAM softwares to convert BMP to relife and generate nc file. BIts is very important. There are many bits, The best bits is ball tip bits. overlap is also important. When it is small, surface is smooth but taking a lot of time.

[ger21]

Originally Posted by DHSYCNC if you choose pinion and rack and want to do 3D carving, surface working piece is not very smooth.

Not true. There are plenty of people using rack and pinion On similar machines with very good results. Just look at ShopBot. All their machines are rack and pinion.

Let the criticism begin!

What ever your gantry sides are made of, they look like a weak point. Are you using hardwood? Whatever it is, it need to be thicker than the 3/4" it appears to be. Especially where the top carriage is bolted to it.

[ivojnob]

What ever your gantry sides are made of, they look like a weak point. Are you using hardwood? Whatever it is, it need to be thicker than the 3/4" it appears to be. Especially where the top carriage is bolted to it.

Thanks for the feedback. Yeah, that's the part that I'm struggling with. The way the pinion has to line up with the rack means I have a particular thickness I can use to join the upper and lower carriages if the R&P drive is going to mount to the outside as shown in the pictures. In this case it's 5/8", so I can either do hardwood, plywood or MDF. I want to thicken the gantry sides above the top carriage as much as possible (like I've shown, but maybe all the way to the inside edge of the top carriage) and bolt the thickened gantry sides to the top of the carriage in addition to the side. I could do 5/8" aluminum for the sides if I have to. I wonder though if that would be stiffer than having the sides be 2-1/2" thick from the top of the carriage to the bottom of the main gantry beam. I still haven't thought much about the connection between the gantry sides and beam.

I was originally thinking of doing something like this older screenshot I made, but I read a comment that said it's better to have the gantry overhang the sides so that the carriage bearings aren't pulled out of line with the 1/4" rails...

[ivojnob]

This picture is a better shot of what I was planning for the gantry sides, although the gantry was just slapped together at this stage.

Notice how the sides have been thickened above the top carriage and extend to the front of the carriage. I still have concerns about the 5/8" thick piece that connects the upper and lower carriages especially with the cutout that I have to make to accommodate the R&P belt drive. If there's any merit at all to this design, maybe I need to consider extending the part between the two carriages further forward so that the hole for the belt drive is in the middle of the part (i.e. the wood is closed at the front instead of the open "birds mouth" that I've shown here)

[SPEEDRE]

May I suggest, that you over build the gantry support beam and laminate it from 5/8 th's AAA baltic birch plywood, with 3/4 in maple substrate. And the side members that connect the beam to the linear unit be wider on those members as, this, is the weakest point of the whole unit. I believe this will help eliminate all the annoying flex and slop of this design.

[ivojnob]

Originally Posted by SPEEDRE May I suggest, that you over build the gantry support beam and laminate it from 5/8 th's AAA baltic birch plywood, with 3/4 in maple substrate. And the side members that connect the beam to the linear unit be wider on those members as, this, is the weakest point of the whole unit. I believe this will help eliminate all the annoying flex and slop of this design.

Not sure I follow. Are you suggesting I make the gantry support beam out of laminations of wood all going in the same direction as opposed to building it like a long, thin box?

[ger21]

The machine I'm currently designing will use torsion box type sides, with the rack and pinion mounting plate mounted inside. Haven't worked all the details out yet. Another option is to reinforce with steel tubing. I'd opt to bolt as well as bond with epoxy, for maximum stiffness.

[DHSYCNC]

Originally Posted by ger21 Not true. There are plenty of people using rack and pinion On similar machines with very good results. Just look at ShopBot. All their machines are rack and pinion. What ever your gantry sides are made of, they look like a weak point. Are you using hardwood? Whatever it is, it need to be thicker than the 3/4" it appears to be. Especially where the top carriage is bolted to it.

======================================================
if module of gear is1.25, rate of reduce 1:2, pulse equivalency is 0.0216mm/step. but we use ball screw, it's 0.0025mm/step. surface is quite different. we had compared two kind of results. if only one result engraving sample, you can't recognize it. On other hand, gear has gap when direction is oppisite, the gap will cause result worse. we had made test by engraving stone, wood, soft and hard wood. more pulse equivalency is small, result is fine.

[freecncplans]

I am kind of confused why you are persistent not using rack and pinions and must use ballscrews? I don't like it when people come in and just use blanket statements like these and assume that something is not worth using when they have no clue.
Sean

Originally Posted by DHSYCNC ====================================================== if module of gear is1.25, rate of reduce 1:2, pulse equivalency is 0.0216mm/step. but we use ball screw, it's 0.0025mm/step. surface is quite different. we had compared two kind of results. if only one result engraving sample, you can't recognize it. On other hand, gear has gap when direction is oppisite, the gap will cause result worse. we had made test by engraving stone, wood, soft and hard wood. more pulse equivalency is small, result is fine.