Rotary SCANNING using your Fourth Axis

For this tutorial, we scan a simple drill bit

Your Fourth Axis is capable of high-resolution rotary scanning, which is usually best utilized when the XZ scan plane lies above the centre-of-rotation of your Fourth Axis. This article will explain how to determine this point, and how to use the Dr Picza software freely included with your MDX machine to capture rotary data.

As an example target for this procedure, we have selected a carbide drill bit to scan. Whilst this target has some rotational symmetry, it is not imperative to have it spin perfectly concentrically in the chuck. If the drill is eccentric when the scan probe senses it, any offset can easily be corrected when the data is shown in its correct polar representation, by redefining the origin.

Calibrating the Y-motion-table position and rotary axis position.

Why do we worry about these two factors?

  1. We want the Y-motion-table positioned correctly before a scan, because the most useful data is obtained when the rotation axis is directly under the XZ scan plane.
  2. Having the rotary axis start at a zero-degree position lets us know absolutely where features are relative to the real-life object. For instance, if you wanted to re-scan a certain area of the object that was still mounted, you know that so long as the Rotary Command reads 'zero' when you initiate the scan, the data will appear in the correct place on your model.

Y-motion-table calibration procedure (in SCAN mode):

If you need a quantitative, numerically repeatable (e.g. for future scans) Y-setback position, the following steps show you how to do this properly through software.

  1. Start by ensuring your interface is set to the Y-AXIS, then turning on your MDX in "scan" mode with a R.A.P.S. sensor attached.

  2. The scan target is mounted in the four-jaw chuck
    At this point, you can mount your scan target in your Fourth Axis after the MDX finishes homing in XYZ. We just used the four-jaw chuck.
  3. Now we can run Dr Picza to manipulate the Y-motion-table. Click "Scanning Area" to bring up the familiar blue/yellow dialogue.
  4. Now enable the "Z Upper Limit" function. Please note, We are NOT trying to find the Z-upper limit, but simply using this as a tool to control the machine until we can find the ideal position of the Y-motion-table for our scan.

  5. We use the "Scan Settings" dialogue to move the machine for now.

    In the case of the Fourth Axis20, we designed it with an ideal Y-setback value of 97.15mm, but each MDX-20 machine is different! You will need to "tickle" a few different spots to find top-dead centre. Please see our Y-setback philosophy article.

    Because the complete Y-distance available to an MDX-20 is 152.4mm, we subtract 97.15mm to give us a close estimation of where to send the Y-axis (about Y=55.3mm from the origin). Now "apply" a Z-upper limit of X=20, Y=55.3. In this case, X=20 was used because the drill's shaft allows us to see neatly from the top of the machine how close we are to its middle.

    You will need to try values closer or further away from Y=55 to experimentally determine the best Y-setback value for YOUR machine!

  6. The probe is ideally exactly above the axis of rotation
    When you you have found the best position of the Y-motion-table for your machine, WAIT UNTIL ALL MOTION HAS STOPPED, and then toggle your interface box to ROTARY mode.
  7. Now do a Z Upper Limit test at X=0, Y=0 to reset the step buffers in your machine. This is important!

Rotary axis position calibration procedure (in SCAN mode):

If you need a calibrated rotation-start position for your Fourth Axis (perhaps you plan on rescanning part of an object, or need to know which way up it will appear in software), you can follow these steps.

  1. Look carefully and as accurately as possible at the Rotary Command readout on your machine. What value can be seen?
    Enter your Fourth Axis calibration number: - - - -

    Enter the Rotary Command value you see: °

    Click "Calculate" and write down this value:      Y = mm
  2. Switch your interface box to MIRROR mode.
  3. Do a Z Upper Limit test at X=0, Y= mm (where Y is the output from the above calculator).
  4. Switch your interface box to NORMAL mode.
  5. Now do a Z Upper Limit test at X=0, Y=0.
  6. Confirm that the Rotary Command display now reads zero degrees. You now have a calibrated start position for your Fourth Axis.
  7. Disable the "Z Upper Limit" function.

Choosing your scan area:

If you wish to scan exactly one revolution, or a fraction thereof, it's helpful to know how many millimeters in the Y-axis will correspond to one revolution on your rotary axis.

  1. How many degrees around your object do you wish to scan from the start point? Use the calculator below to help determine which scan area values to use.
    Enter your Fourth Axis calibration number: - - - -

    Enter your scan range (not over 360 degrees): °

    Click "Calculate" and write down this value:      Y2 = mm

  2. The scanned data as seen in Dr Picza.
    In Dr Picza, open the "Scan Settings" dialogue, and consider what X-values you want to scan between. You won't want the needle to hit the chuck, and you won't need to scan air / empty space beyond where your object lies. Your Y-values should extend between Y1=0mm and Y2=mm , (Y2 is the output from the above calculator).
  3. Set your X-scan pitch, Y-scan pitch, Z-Bottom and Z-Upper-Limit values. We advise using the finest Y-scan pitch possible, to fully utilize the resolution available in your rotary.

  4. The scanned data as seen in Dr Picza.
    Do an Area Test to confirm your X and Y (Rotary) ranges are set correctly. If you've chosen to scan a full revolution, the needle should "tickle" points at 0°, 180°, and '360°' on your rotary axis.
  5. Hit 'SCAN', and away your machine will go! This is the data displayed in Dr Picza after we scanned this test target.
  6. Make sure you SAVE your data in the native Pizca .PIX format; you will likely be converting the file later.

"That's doesn't look like what I scanned!"

This is the polar representation of the scanned data.

Actually, the data you see is a complete and accurate record of the surface of your target. To represent your data accurately involves a bit of discussion and a few more steps. Please now move on to our Picza-to-Rhino conversion tutorial and our Rotary data in Rhino (representing your model) tutorial, to get your data looking like this!

Inquiry Form