Machine Rigidity

[JerryBurks]

I am planning to upgrade my new Pro Plus with supported guides like shown in an earlier post. For the precise parts I want to machine the stiffness of the Shark is not sufficient, especially at higher feed rates.

Before I start taking everything apart, I measured how rigid the machine actually is as shipped. The most compliant points I found are in the center of the table and at maximum X-distance. I thought others may be interested and here are the numbers (measured values are 1/1000" per pound load in this direction). Considering my makeshift measuring setup I would not expect it to be more precise than +/- 20%

Center Table
X-deflection 2.7
y-deflection 3.3
z-deflection 1.1

Max-X, Center-Y
X-deflection 5.6
y-deflection 3.6
z-deflection 2.7

Interestingly, the machine gives more when at the ends of the x-axis than in the center. I did not expect that.

Once I am through with the modification (may take a few weeks) I will post an update.

JB

[CountryWoodCrafts]

Cool guy . . . I will be looking forward to your mods and measurements . . .I changed my 62" top which was two MDF pieces to 3/4" Oak Plywood and removed a lot of bounce.

[FixitMike]

The increased deflection in the Z direction when the head is at the ends of the X travel could be explained by deflection in the Y guides. With the head closer to one of the guides, the vertical forces will be concentrated on that guide, instead of being shared by the guides on both sides, thus more deflection. This also explains why the deflection in the Y direction does not change significantly with X movement. (Although it does not explain the change in the X direction.)

This is similar to the conclusion that we arrived at when measuring table flatness. The measured upward "bulge" in the center of the table can be explained by sag in the Y-axis guides when the gantry is near Y center.

If the main source of deflection is in the Y guides, then there is not much that can be done to improve the stiffness without substantial rework, such as replacing the existing guides and bushings with ones of a larger diameter.

[rungemach]

Based on Jerry's experiment, I did some quick measurements on my modified shark pro, and came up with this:

Method:
router position at center of table x and y with bottom plate of router mount assembly even with bottom plate of z axis "bearing box".
A 1/4 inch dia rod extending 1.5 inches down from bottom plate of router mount. (to simulate a 1/4 inch straight router bit)
16 ounces of force applied to the end of the rod (router bit) in the x y and z directions individually (simulates cutting force)(used 1.5 lb high quality spring scale)

Measured the movement of the router tip using a dial indicator mounted to the cutting table.
Measured the movement at the z axis bearing box to eliminate the flex contribution of just the z axis (as that has not been modified from stock yet, but is next on my list)

The dial indicator is graduated in .001, the tenths are estimated visually.

x direction router tip movement with 16 oz x force .0018" movement at the bearing box .0008"
y direction router tip movement with 16 oz y force .002" movement at the bearing box .0005"
z direction router tip movement with 16 ox z force .001" movement at the bearing box .0001"

This showed me how much flex is still left in the z axis and router mounting plate.
The machine base and gantry have firmed up pretty well with the modifications I have done already and I am happy with their measurements.
The modifications were mainly replacement of the unsupported rod bearings with supported rod bearings and a few other things.
I can go into detail if anyone is interested.

These measurements do not include backlash as the force is applied in one direction only, then released.

These results will get me going on beefing up the z axis/ router mount assembly sooner than I thought.
It looks like over half of my remaining flex is still left in the z bearings and router mount assembly.
The worst is the y direction as could be anticipated. Looks like a stronger router mount assembly is in my near future.

I hope this info is helpful.

[JerryBurks]

..........If the main source of deflection is in the Y guides, then there is not much that can be done to improve the stiffness without substantial rework, such as replacing the existing guides and bushings with ones of a larger diameter...........

I believe I found a way of stiffening this up without major reconstruction. I am planning to widen the table with another blue aluminum slat, provide serious square tube reinforcement across the slats and mount the new 37" supported guide rods on top of the table edge, next to the gantry vertical plates. Then I can mount the sliding blocks with an aluminum angle bracket to those plates. This will not cost me any x-movement of the router bit but it will limit the maximum width of the blanks a little. The major advantage will be no change to the undercarriage and the guides have a very solid and parallel connection to the table.

Since I am planning to use a small spindle I will throw out the router mounting bracket completely. Actually, I may machine a new bracket for the spindle before I take it apart.

Looks like a fun project to me. I can not buy anything and just leave it alone. Bad and expensive habit.

[rungemach]

Just a quick update on my shark's flex measurements.

I finished an aluminum router mount assembly to replace the plastic factory unit.
The new one improves rigidity of the unit, and the new flex measurements are as follows.
Measured at router bit tip with 1 lb of force, router centered on tabletop.

X direction 0.0010
Y direction 0.0007
z direction 0.0005

I'm currently running a Makita router head, and made mounts for my Colt and Dewalt while I was at it.
the Makita has a good deal of cooling air that blows directly past the bit, and keeps the cut area free of dust and chips very well.