Why I got this strange side finish with my router?

Hello all,

I decided to mill some simple shapes from a 10mm thick alu-plate. Before I always cut max 5mm alu-plates without problems.
So I cut the designed holes and then I used some screw to firmly fix the stock at the MDF fixture. Finally, I milled the countour.

Here the problem: only the bottom side came out with a shiny finish. All other sides have a matt surface in the upper part, and a shiny surface only on the bottom zone:

https://i.ibb.co/1rMBv1K/sideplate.jpg

Why?
I immediately told to myself: "the spindle is not perfectly vertical".
But this explaination doesn't work, the spindle cannot be angled in three different directions at the same time.

Here below the strategy I used (4 mm dia, 1 flute end-mill, DLC):

https://i.ibb.co/WpC2rvx/strategy.jpg

I'm not an expert, but it seem quite ok to me. The first and the second cut was made by using an helix-ramp with a "DOC" of just 0.8 mm.
All of the three cuts was made as climbing cuts. No chattering at all.


The manufacturer of the machine declare a precision/repeatability of 0.05 mm (i.e. around 2 thou), and I can only confirm that after one year of continuos use.

I'm scraping my head... What is the cause here? Lack of precision? Spindle run-out? Or what?

I'm pretty sure that if I leave more stock material before the final spring-cut all, then all of the surfaces will came out shiny and nice. Nevertheless, I would understand what's the matter here.

Any idea?

Tool is dull on the bottom maybe? Not enough coolant?

... Green edge had minimal stock, where red edges a wider width of cut ?
... How much finishing allowance was left after the roughing(step-down) passes ?

You need an amount that would clean up any toolflex that may occur.
The finishing pass needs a small amount to actually cut... rubbing is actually making your tools blunt.

Quote:

---

Originally Posted by Jim Dawson

Tool is dull on the bottom maybe? Not enough coolant?

---

+1

Quote:

---

Originally Posted by Jim Dawson

Tool is dull on the bottom maybe? Not enough coolant?

---

If so, why the south side is shiny and nice, while the east/north/west sides have this crappy finish on the top side?
The tool is rotating on his axis, so its irregularity should be applied on all of the sides of the plate in the same manner... or not?

I cannot use coolant at the moment, for this reason I'm using Diamond coated tools (DLC or TaC). This has worked very (very) well for alu-sheet up to 6 mm of thickness.

I lowered the feedrate and the DOC to reduce the amount of heat generated.

Quote:

---

Originally Posted by Superman

... Green edge had minimal stock, where red edges a wider width of cut ?
... How much finishing allowance was left after the roughing(step-down) passes ?

---

All of the sides have the same WOC.
After the roughing pass I left 0.1 mm on all sides of the plate. Then I applied the spring-cut on all the countour in one pass (i.e. DOC = 10 mm).

I also thought that this was because of poor precision/repeatability of the router: i.e. the tool is moved at a slightly translated position at the beginning of the finishing pass.
But, again, this explaination doesn't hold because all of the west/north/south sides have this poor finish in the top zone.

Quote:

---

Originally Posted by Superman

You need an amount that would clean up any toolflex that may occur.

---

Yeah. This is one test that I want make.
I will double the amount of material left after the roughing pass (from 0.1 to 0.2 mm - that is from 4 to 8 thou circa).

Quote:

---

Originally Posted by Superman

The finishing pass needs a small amount to actually cut... rubbing is actually making your tools blunt.

---

Yeah, you're right. Rubbing instead of cutting is something to avoid.
Nevertheless, only on the south side of the plate, very nice chips was jumping out. Shiny and long. This means that the cutter was very sharp and not rubbing.
On the west/north/east side was a different story...

Unfortunately, I was short on time and I milled all the pieces in the same manner (good finish was not important for my customer).
Now I ordered more material to make further tests.

My tests will be:

1. make exactly the same cut with a more rigid 6mm cutter
2. leave more stock material before the finishing pass

Any other test I could make?

What is the cutting flute length?

If the flute / cutting length is shorter than the material depth, you will get rubbing and it will look like that
(Standard cutting bits do not have a relieved shank. You can get relieved shank bits that have a small amount ground off the shank diameter to prevent this.)

I would suggest that the finish pass be tested with a fresh, new tool....
This would eliminate any thoughts that the clearing & roughing passes have slightly dulled the cutting edge, & it may be the dulled cutter edges that create the "shiny & nice" surfaces.

I myself use 0.2mm as a finish allowance. Where 0.1mm is still n OK amount.... I assume you are roughing & finishing with the same tool.
Word of advice... if you do larger quantities, separate the tools so that the rougher is replaced when worn. And the finisher gives the same size results and consistent finish

Quote:

---

Originally Posted by gimpo

All of the sides have the same WOC.
After the roughing pass I left 0.1 mm on all sides of the plate. Then I applied the spring-cut on all the countour in one pass (i.e. DOC = 10 mm).

I also thought that this was because of poor precision/repeatability of the router: i.e. the tool is moved at a slightly translated position at the beginning of the finishing pass.
But, again, this explaination doesn't hold because all of the west/north/south sides have this poor finish in the top zone.



Yeah. This is one test that I want make.
I will double the amount of material left after the roughing pass (from 0.1 to 0.2 mm - that is from 4 to 8 thou circa).



Yeah, you're right. Rubbing instead of cutting is something to avoid.
Nevertheless, only on the south side of the plate, very nice chips was jumping out. Shiny and long. This means that the cutter was very sharp and not rubbing.
On the west/north/east side was a different story...

Unfortunately, I was short on time and I milled all the pieces in the same manner (good finish was not important for my customer).
Now I ordered more material to make further tests.

My tests will be:

1. make exactly the same cut with a more rigid 6mm cutter
2. leave more stock material before the finishing pass

Any other test I could make?

---

You may have some backlash in those moves, as well as tool flex or Z axis flex, what spindle are you using

Quote:

---

Originally Posted by pippin88

What is the cutting flute length?

---

Tool LOC is 12 mm. More than enough, IMO.

Quote:

---

Originally Posted by Superman

I would suggest that the finish pass be tested with a fresh, new tool....

---

That was my new tool... I just tested it before with a finish spring cut long 3 or 4 centimeters on a raw block of aluminum. And it was cutting very well, nice chips.

Quote:

---

Originally Posted by Superman

I myself use 0.2mm as a finish allowance. Where 0.1mm is still n OK amount.... I assume you are roughing & finishing with the same tool.

---

Yes, I will try to increment the finish thickeness to 0.2 mm, and I'm sure the problen will disappear. At the same time I'm curious to understand from where this "error" comes.

Quote:

---

Originally Posted by Superman

Word of advice... if you do larger quantities, separate the tools so that the rougher is replaced when worn. And the finisher gives the same size results and consistent finish

---

That's a good idea! Thanks!

Quote:

---

Originally Posted by mactec54

You may have some backlash in those moves, as well as tool flex or Z axis flex, what spindle are you using

---

Absolutely. Backlash could be the source of this behaviour. But why the west/north/east sides are affected and south one is not?

If I remember correctly, the cut started at the south/west corner, and I applied a climbing cut in a clockwise manner. So the cutting path was:

1. positioning at the south/west corner,
2. west side,
3. north side,
4. east side,
4. south side.
5. stop and raise up the tool.

(The south side is the only one with a nice shiny finish.)

I have no many details about the spindle: it is the usual anonymous 2.2 Kw spindle from China... Run-out could justify this results in your opinion?


P.S.
I was in hospital last night because of blood pressure problems. Now I have to stay at rest for two or three days. I will make some (definitive?) tests as soon as possible.

I have a feeling that an hour or so of checking for backlash at various points within the work envelope would be a smart move.Look for signs of wear on the ballscrews and guide rails.All it takes is a dial gauge and a bit of pushing.Don't overlook the effect of cumulative wear on say the gantry rails and the Z axis where each has a little wear and when combined,it becomes a greater amount.The other old standby is to cut two pieces of the same material to the same shape and to use a climb cut on one and a conventional cut on the other so that the sizes and the surface quality can be compared.