A plasma table is a strange environment for a gear rack. It puts a hardened steel tooth pair into a stream of abrasive, electrically conductive dust, runs it for hours at a time at high feed rates, and then judges the result by whether the edge of a 12 mm plate came out clean. No other rack application combines those three things.
The selection problem is also different from what people expect. On a router you are usually chasing position accuracy. On a plasma table, position accuracy is rarely the limiting factor - the kerf is wider than most of the errors you are worried about. What matters is repeatability, edge quality at speed, and staying alive in the dust.
Here is how we work through a plasma rack selection, and where the decisions actually land.
Start from the cutting force, not the table size
The first number we ask for is not the table length. It is the maximum cutting force the gantry has to push through, which comes from the material thickness and how fast the operator wants to cut it.
That force, applied at the pinion radius, is what sets the tooth load, and the tooth load sets the minimum module. Our racks run from module 1.5 up to 12, but the useful band for a plasma gantry is narrower than that. As a rough guide on the machines we supply:
| Module | Typical plasma gantry | Comment |
|---|---|---|
| 1.5 - 2 | Light tables, thin sheet, high acceleration on a small gantry | Finer positioning resolution for the same pinion and encoder |
| 3 - 4 | The common choice - this band carries the thrust of plasma and laser gantries | Where most of our plasma customers land |
| 5 and above | Heavy plate, thick material, long spans | Check the resolution trade-off before going up |
And that trade-off is the part people miss. A bigger module means bigger teeth, which means the pinion moves the axis further per motor revolution - so for the same pinion and the same encoder, a higher module gives you coarser positioning resolution. Going up a module to gain strength costs you resolution. Pick the smallest module that actually carries your cutting force, and solve the rest with the drive.
If module, tooth form and section length are all still open, we run through that decision separately in how to choose a gear rack for CNC - worth reading before you ask for quotes, because it is the difference between comparing suppliers and comparing specifications.
Straight or helical teeth
Straight racks are cheaper, simpler, and on a small table they are entirely adequate. Helical racks engage more than one tooth at a time, which gives smoother motion, less noise, and higher load capacity - at a higher price.
For a plasma gantry that has to produce clean edges at high feed rates, helical is usually worth the premium. The reason is not noise, although that is a bonus. It is that the load transfer is spread across several teeth at any instant, so the small pitch variations between teeth are averaged out rather than passed straight into the cut. On a machine cutting at 5-8 m/min the difference shows up in the edge finish, particularly on corners where the axis is decelerating.
If your table only ever cuts thin sheet at moderate speed on a short span, straight teeth will not embarrass you.
Whichever you choose, order the rack and the pinion as a pair. Module is the obvious match, but the pressure angle has to line up as well, and a pinion that is one pressure angle out from its rack will run - just noisily and with an odd wear pattern that takes a few months to become obvious. It is worth writing the pressure angle on the order rather than assuming it will be matched.
Backlash, and why it shows up on corners
Rack and pinion backlash is set at assembly and then drifts as the mesh wears. There is no way around that on a rack drive - unlike a ball screw, the mesh geometry changes as metal is removed.
On a plasma table, the visible consequence of too much backlash is a small witness mark just after every direction change, because the pinion crosses the backlash gap before it starts pushing again. You will see it on the corner of a rectangular cut, or on the tip of a sharp detail. Too little backlash is worse: the mesh runs hot, noisy, and the teeth load on both flanks at once.
Two practical notes. First, set backlash with the gantry at mid-travel if the rack runs the full length, because thermal growth over a long axis changes the mesh between cold morning and mid-afternoon. Second, re-check it after the first hundred hours - a new rack and pinion beds in, and the setting you had on day one will have moved.
The dust is the real enemy
Plasma cutting produces a fine, abrasive, conductive dust, and it settles on everything. On a rack drive it does two things. It works its way into the mesh as a grinding paste, and if it reaches the motor or driver it can short things out.
Lubrication is the defence, and the interval on a plasma table is shorter than the general figure. Our standard guidance is to grease the rack teeth and pinion every 100-200 operating hours with a lithium-based or gear grease; on a dusty cutting environment you should halve that. On a table running two shifts, that means greasing weekly, and it is the single maintenance task that most affects how long the rack lasts.
Watch the joints. An unlubricated rack wears visibly at the joint sections first, so those are the places to inspect if you want an early warning. Grease the tooth flanks properly rather than spraying and hoping - both flanks, because a gantry decelerating a heavy torch carriage loads the non-driving side regularly.
The other half of the answer is separating the dust from the drive. If the rack sits in the open under the cutting bed, expect to grease more and inspect more. Where the design allows, a brush wiper on the pinion mount and a cover over the rack on the underside of the beam pay for themselves quickly.
Long tables and joins
A 6 m plasma table is normal, and that means several rack sections joined end to end, with the gantry crossing every joint hundreds of times a minute at feed rate.
The joints are where the trouble goes, and the fix is in the setup rather than the parts. Sections should be supplied with machined end faces so the joint closes metal to metal, and the second section should be shimmed until the tooth line is continuous. Check it with a dial gauge riding along a tooth across the joint - a straight edge along the back of the rack will not show you a height step, and a height step is what leaves a ripple in the cut surface every time the gantry passes.
Suppliers sometimes include a joining pin to locate sections axially. That pin does not remove a height step. It is worth knowing that before you skip the shimming.
How the rack fits with the rest of the table
The rack is one link in a chain that includes the controller, the plasma controller, the torch height control and the drive.
Height control matters more than people expect in a rack selection conversation, because the torch carriage is moving weight up and down while the gantry is accelerating sideways. That combined load is what the pinion sees, and it is why a table that looks light on paper can need the top of the module 3-4 band.
The controller side has its own requirements - automatic corner speed control and a height controller interface are the two features that most affect cut quality on real work, and both interact with how the axis is tuned. A rack that is set up well will let the controller do its job; a rack with a step at a joint or excessive backlash will have the controller chasing an error it cannot fix.
Our racks are S45C or 42CrMo, quenched to HRC48-60, with four-side and tooth-flank grinding holding DIN6-8 and a tooth error around ±0.01 mm/m. For a plasma gantry that is more accuracy than the process needs - the reason to buy ground rack here is not the tolerance on the cut, it is tooth-to-tooth consistency, which is what keeps the motion smooth at speed.
If you are specifying rack for a plasma or laser table, the four numbers that let us answer properly are: table length and travel, maximum material thickness, the cutting speed you want on that thickness, and the pinion you already have or intend to use.
We supply module racks with matched pinions for gantry routers and plasma tables, and we would rather size it to the axis than to the catalogue. Send the details through here and we will come back with a module, a tooth form and a pinion recommendation - usually the same working day.
