Arranging project cuts for accuracy

[Paul McGrath]

I just posted the customer support regarding this, but I thought would ask for advice on the forum as well. We are currently cutting 12mmx36cmx121cm corian and having accuracy problems on a double sided project.

Given the accuracy of the machine is dependent on the number of rotations the tracking sensor makes(board travel distance) and the initial edge sensor accuracy:

1) Can we change the order of the cutting so the cuts needing the most x-accuracy comes first? Or before error occurs.
2) At what point is it better to break up a project into multiple runs?
3) At what point is it better to agglomerate cuts into a pattern using STL and raster cut?
4) Is the x-accuracy drift of the raster cuts accumulative on the passes or global? Should we do a path cut around the outside and hog out the material using the raster?
5) What location on the board is the most accurate on the front and the back (I assumed the 0,0 point but from the recent cut results I'm less sure).

[DickB]

Are you using a sled?

[Paul McGrath]

No sled, the pattern is cut with 3.5" excess material on either end and about 3/16" untouched material on the edges to act as a rail.

[dbfletcher]

No sled, the pattern is cut with 3.5" excess material on either end and about 3/16" untouched material on the edges to act as a rail.

Can you post a pic of your completed carving so we can get an idea of what you are talking about for inaccuracies? Generally my machine it typically with in a few thousands which is more than enought accuracy for what I do.

[Digitalwoodshop]

X is all about the Brass Roller....

IF I were doing your project I would get some Transfer Tape used by Vinyl Sign people and apply vinyl that looks like masking tape. Apply it to the bottom of the Corian to let the sand paper belt get some bite. And a strip of masking tape too.... That will let you get better accuracy in the X.... The Brass Roller is slipping.....

ALSO since the feed rate of the machine is setup for WOOD and you are doing Corian... What will happen if the spin of the bit and the feed rate does not cut away the material fast enough then something will give.... The X will slip or the Y will click and skip a tooth.... I had it happen when my bit became dull cutting 1/8 inch Fiber Re Enforced Plastic. Got the Y Clicking...

The SOLUTION.... If you do any Vector Cuts with the 1/8 or 1/16 inch Cutting or End Mill bit... Change the bit selection in designer to 1/4 inch ball nose BUT load the 1/8 or 1/16th.... The advantage is that when you assign the 1/4 inch bit it will leave the feed rate in 1st gear..... A slow feed rate.... A normal feed rate would be 2 speeds... 1st on corners and 2nd gear on the straight areas.

And yes, hogging out the center in Raster and clean up the perimeter with a Vector Cut and depth.

Just know the machine was not designed for Corian. The feed rate is too fast and we have NO control over the selection of what cuts first.... Normally Raster then Vector... But no choice of a few vector cuts on a project.

Doing 2 projects is a option... Using Place on END is better as you don't have the measuring twice as in Center on board....... Just once for place on end... Do a search using the "place on end" with picture I have posted. I place a 4 inch dead area in my Designer on the right side and have a 4 inch blank area on the board.... I tape blocks on the end... You could make a sled with a slot to drop in a Corian piece....

Remember Masking Tape on the bottom of the board gives the brass roller something to dig into... Like a Rack and Pinion...


Good Luck,

AL

[earlyrider]

Keep in mind, the machine doesn't really like thicknesses even slightly less than 13mm (or 1/2 inch). This might be part of the measurement problem.
Ron

[Paul McGrath]

Fri-07-Jan_01:36:27 PM

Thanks for all the replies and sharing your suggestions. I didn't think of changing the bit size to reduce the feed rate also 4" sounds better than 3.5" on the end. Thanks for the heads up on the FRP. We were planing on using the carvewright to cut some samples. Still disappointed that we have no control over the order of operations within a project. However I am leaning towards breaking projects up into smaller chunks or rough cutting and then doing final cuts.

I have attached a photo of the miss cut on one of the components. The odd thing regarding this is the region is the cut longer than in the designer by about 1_cm. I can't think of any hardware related issue that will give this problem except slipping of the tracking roller. Even so I will go into details on how this board was setup. Sorry for the procedural tone in the writing, trying to hit two birds with one stone.

We did some pre-cuts using MDF to make sure the design was what we wanted. We did discover some setup refinements during the process that produced a final error free cut in MDF.

Setup steps done for each component:
1) Temperature above 12_C [was between 18-20_C];
2) Make sure the board is straight and within -0+1_mm of the board dimensions in designer.
3) Draw a line with permanent marker on the far edge face (used in later steps);
4) Check the head pressure with the floor scale (85-95_lb);
5) Run through a sensor check;
6) Put duct tape on the bottom at the edges to help grip the board and to allow good contact with the tracking roller. [Afterwards witnessed the tracking roller teeth marks all the way along the length];
7) Placed the board in the machine on the LHS and lined up the end of the board just above the tracking roller Adjusted the elevations of the external roller stands so the board just contacts the tracking roller;
Repeat above with the board on the RHS of the tracking roller;
9) Clean the edge sensor with a cloth and make sure the reading is in the 135+ range;
10) Firmly push the board onto the machines fence. Use the 3/8" Joining bit and run the join command until the permanent black marker is totally off and there is no drift between the fence and the board edge.
11) Vacuum up shavings and again clean the edge sensor with a cloth and make sure the reading is in the 135+ range;
12) Run the measure board program. The measurements should be within +-1_mm.
13) Run program with: No to keep rollers on, Original size, On corner.

Some additional comments about the Carvewright. It's a Model B with rock chuck and rubber belts. There is dust collection system that is based on the pattern in the pattern shop and the forums. It collects most of the air born free dust and prevents a large collection inside the machine, but it is not strong enough to pull the dust out of the cuts. We don't vacuum during the bit changes because we unfortunately found the static electricity the machine resets.

During the process of cutting the corrian the only error (tracking) was caused by the measuring the length of the board. It happens when the uncompressed roller tries to move onto the corian board and it stops moving in the x-direction. We didn't have that problem with the MDF and think it is the sharpness of the edge. When this happens we gently aided the x-drive with a small push and the compression roller moves on the corian and give an accurate length. In the future we will file down the edge of the boards to see if this prevents this. For the rest of the project the roller remains on the board so this only happens in the beginning.