Machining 'scribe' and 'ruler' lines

[PCS74]

Hello all,

I am designing a large jig table and need some advice on machining. We have an 8ft x 12ft table(1 inch thickness). We want to run 'scribe' lines on a 2 inch grid pattern and 'ruler' markings on all four sides. No particular constraints on scribe line size or ruler scale has been set yet, but the smaller the better in both cases.

I am trying to look at this from the machinists side of the fence, so my question is what would be my best options as far as processes (machining, laser, other)? What would be my limiting factors? What types of pitfalls can I expect? What might you do for something like this?

I am open to any and all suggestions, recommendations and criticisms, and the more input the better.


Chris S

[mc_n_g]

A lot depends upon the purpose of the jig and table. What do you expect out of it? Your ability to place something on the table at an exact coordinate will be limited greatly. You can scribe lines down to the .001" but what level of accuracy is needed? 8' x 12' by 1" (mystery material) is large for most people to machine (and also expensive). You really need to evaluate what you are doing and why? What is the ultimate goal for what you are working on? (rhetorical question)
Will a good T-square and Rapidograph pen work? How will the stops for the jig be embedded in the table? How accurately can you place those wither with bolts or welding? Each process adds error into the overall outcome.
If you think you are going to get .001 accuracy over an 8' table there is too much involved. Look at your process and what you need first.

[PCS74]

Table will be steel. We are starting with a 1.5in. thick plate and will end up around 1in. We are reinforcing with a 0.625in. by 12in. rib system. For the rib system think of a rectangular box with an X and a + drawn inside.

We are placing 3/8-24 threaded holes on a 4in. x 4in. grid pattern. Each hole will be counterbored for a 5/8in. dowel and chamfered. The scribe lines will be placed on a 2in. x 2in. grid pattern. Each edge of the plate will have ruler marks on 1/16 in. increments.

The plate will be used for the reverse engineering of automobile sub structures and bodies, the design/setup/evaluation of tooling, and positioning /construction of prototype vehicles. It needs to work well for highly accurate representations all the way to quick and dirty measuring. We will be using it for laser scanning and coordinate measuring purposes initially, but it's use will progress from there to the others.

As for accuracy, no I do not expect 0.001in. But, we have been told we can get +/- 0.005in. on parallelism and flatness from grinding (at significant cost) and comparable or better from fly cutting (at less cost). Surface finish is important, but we don't need anything like a 16 or 32 u-in., more like 60 to 100 u-in. or even a little rougher, so if the machining tolerances are good we are ok there.

Hole and grid positioning is a different story. Lots of holes and lots of lines. I'd like +/- 5 thou. on location (from hole to hole) for these, over the entire table, but I don't know if that is possible with such a large envelope(am I dreaming here?). I've done some general machining in the past, but not enough to know what good machines and good machinists are capable of these days.

Obviously, the better the table is the better the final product is, but I don't write the checks so it's quality vs. cost as always. Comments?

Tooling for all these holes is another story. The machinist recommends building a custom tool that will perform all four operations drill, tap, counter bore for dowel and chamfer. Sounds like it will save time and improve overall accuracy. I would like to hear your comments here as well.



Well, there is some more info, I would love more input from anyone.

thanks,

Chris S

[Geof]

Some comments:

If you start with 1.5" plate and machine it down to 1" by taking material off only one side you may encounter serious warpage. You mention stiffeners, presumably welded, on the backside but your description suggests there will be largish areas between the stiffeners. It is possible you could find the top surface after machining tends to go convex in the areas between the stiffeners by more than your hoped for 0.005".

But perhaps you plan on normalizing or stress relieving the whole thing after fabrication in which case no problem.

Regarding the threaded holes/counterbores I would suggest you get the counterbores bored separately from the tapping hole. It does not matter if your tapping hole wanders a few thou but you do want you counterbore to be accurately positioned.

[chop]

Instead of trying to reinvent the wheel you may want to look into a ready made solution. I know Bluco has a large modular plate system with accurately spaced holes and scribed lines. You may find it would be cheaper to buy something like that instead of attempting to make it. Plus you have the added benifit of using all of their fixturing.

[PCS74]

We looked into Bluco along with some others. After getting a bunch of quotes we did our homework. We drew up some initial ideas and got quotes on materials/machining/etc. We can build a table that will fit our needs much better, be much larger, and have comparable or better tolerances than most for about half the cost.


Chris S

[DareBee]

You should have no problem getting the hole to hole at +/-.005.
Anyone with a boring mill even half that size can use the first quadrant of holes to reposition for the next quadrant.
Sure, it will be somewhat fussy, but this is quite normal for those of us in custom machining.
There are also shops out there that have massive gantry mills that can do it in 1 shot. You mentioned you got a price on milling the surface, I would assume thier machine is big enough.

I would recommend that you put hardened steel sleeves in for the dowel counterbores (drill bushings work well).

[Caprirs]

It might be so obvious that it doesn't get mentioned, but ensuring that the plate will be manufactured at the same temperature as your final installtion is important on something that size. For steel, you'll have approximately 6 millionths per linear inch per degree F. That means if the plate is manufactured at 65*F and your final installation is 75*F, the long dimension (12 ft) will grow nearly .010".

There are plenty of big gantry mills which could do the entire plate in one setup which would assure the precision of the hole locations and scribe/ruler lines. I do not know how many of those machines are in climate controlled facilities. Before commiting to the job, you may ask the owner for records of the most recent calibration.

Laser would be ideal for the scribe/ruler lines but not so great at the holes. Same goes for water jet. Both processes tend to be +/-.005" in the hands of skilled operators, worse if the machines are not well maintained. Also, the lasers and waterjets tend to create tapered holes which means you'd have to finish machine the holes with a seperate process.

[PCS74]

DareBee and Caprirs,

Thank you very much for your input. Hardened inserts is a great idea and I had not even considered temperature dependence(not so obvious on something like this I guess).

Chris S

[Unabiker]

In the several years I spent building 40% scale models of race cars used in wind tunnel testing, I used a variety of surface plates. We had one plate that was close to the size you are talking about (8'x12"), two that were 4'x8' and 4 that were 3'x6'. These plates were used for a variety of tasks, from reverse engineering parts of cars to building entire models to a very high level of accuracy.
A vital feature on these plates was a line scribed down the center of each plate. This allowed us to center up the models on the plate and gave us someplace universal to reference measurements from. I think a 2" grid would have been overkill, but we would frequently tape long rulers down to along the sides of the plates parallel to the scribed centerline. For the rulers you are suggesting along the sides of your plate, it may be beneficial to include a scale in millimeters, as a large portion of vehicles are designed in mm. It would also be to your advantage to have the outer edges of your plate machined square so that you could use a large T-square to assist with your measurements.
None of our plates had bolt holes or dowel pin holes. We did use a lot of magnet bases for holding things securely to the plates. All of the laser scanning was done with a Faro arm with a laser head that sat on a tripod off to the side of the plate.
For the most part, it seemed to be best to have about a foot of room all the way around whatever you were working on. Less than that, you would be knocking things over. More than that would tend to collect extra stuff. The large plate was always collecting stuff, no matter how many times you cleaned it off. Nature has a way of filling a vacuum.
You will find that it is important to have coasters readily available near plates like this. It is amazing how fast the heat can get sucked out of a cup of coffee sitting on a large steel plate.