Thread: Machining Brass problem

Hi, I am no expert but your end/slot mill bit looks to be about 6mm in diameter. According to a table I have, brass is slot milled at 200-400 surface feet per minute (sfpm). Taking a mid point of 300 sfpm, that translates to about 4,800 rpm. Brass is usually machined dry.

Hi Blu_Rock,

Thanks for the reply. My bit is actually 2mm, so this would mean a higher spindle speed. I am running it at 6000rpm. The chart your using, is this available on line for reference.

I seem to have got around the issue by running the job twice and using kero as lubricant. The second pass seems to get the edge nice. Unfortunatly I can't run my spindle less than 6000 rpm on the larger machine, unlike my mill, but then my mill can't machine a piece this large. Cought between a rock and another rock.

Thanks for you feedback, greatly appreciated

/M

Hi Mike,

Firstly, I must say that i have never used a CNC router/milling machine. I was surprised to hear that the cutter was only 2mm diameter. Thus I must ask

1) Why not use a slot cutter followed by the largest diameter end mill possible and shorter in length? - less vibration/chatter!!

2) How do you clamp it down? - double sided tape!!

3) How thick is the plate you are using for this task?

Have the body of the cutter(s) as far into the collet/chuck as possible.

Regards
MH

Hi MH,

1/ I could get away with a 2.5mm cutter for the tightest portion of the job, but i don't have a 2.5 handy.What I should have done was have a couple of tools for the job. the larger one could have performed my clearance cuts and the 2m, the fine cuts in the tight spots. I need to try this before I start on an expensive piece of brass, so next on the agenda is another wood run with this type of setup. The pain with my CNC is I have to change out the cutter, reset the height and continue. I will see how this goes with the wood and try to learn where problems can occur. One thing with changing bits is that there is always a little "runout/accuracy issue" with the machine and i get ridge lines on the edges. Once again I can get rid of this by doing a light cut at the end, which I will also try setting up.

2/ Picture 5 shows my block bracing. Not the best as i found there was a very slight discrepence between one side and the other. On the agenda is to make a vacuum table for the cutting area of the machine. This should sort the level issue and clamping problems.

3/ 2mm - This was all the way up. I broke a couple of cutters in the process which was due to sloppy setup. Not checking the correct feedrates were entered, etc.

I hope this shed some light on what i did.

PS: got around the problem for this job by running the process twice. The 1st run got it close and the second cleaned it all up nicely. Used kero for cooling and finish.

/M

I know nothing about CNC machines etc. , but I wonder if the problem you are having is the same as the problems encountered drilling brass with a standard twist drill. Machinists used to keep a set of straight fluted drills (negative rake angle?) just for drilling brass. Standard drills dig in and cause all sorts of problems.

Just a thought.

Originally Posted by Avery

I know nothing about CNC machines etc. , but I wonder if the problem you are having is the same as the problems encountered drilling brass with a standard twist drill. Machinists used to keep a set of straight fluted drills (negative rake angle?) just for drilling brass. Standard drills dig in and cause all sorts of problems.

Just a thought.

Hi Avery,

Very true. The same type of drills (negative cutting edge) are required for acrylic (and some other plastics) otherwise they "grab" and pull down through the material rather than cut it.

Hi Seafurymike

Glad to see you appear to have cracked it and come up with a solution. Generally brass is ok without coolant in some set ups. Using a vacuum to suck up the chips would be a good idea as well those little chips may impede on getting a good finish as well.

Regards
MH

Hi all,

depending on the percentage of copper and zinc in the brass alloy, it can either be a bit more on the tensile (like steel) or on the brittle side (like cast iron). Brittle stuff has a tendency to be left with frayed edges (tiny broken off bits) and tensile stuff has a tendency to show burring (which is in fact deformation due to unsharp tools or forced forward working speed; the material was pushed out of shape rather than cut in a clean way). Have a look at Wikipedia's article on brass, to see how many additives brass can have in order to make it moe corrosion resistant or to give it certain mechanical properties. Such additives to the alloy can have repercussions on the ways of machining.

Tables can be helpful, but common sense can also predict a lot. Your bit has two cutting edges, meaning it cuts 12,000 times per minute ( = 200 cuts per second) at 6,000 rpm and 24,000 times (= 400 cuts per second) at 12,000 rpm. You spread out these cuts over 125 mms forward working speed per second, meaning there is one cutting movement at every 0.8 mm. That's not a lot for a tool diameter of only 2 mms; you can't expect a smooth clean edge from that at such forward speed.

Your problem can be seen as a mathematical equation, in which there are some given facts (that can't be changed) and some bits which have to give (variables that need to yield a bit in order to make the equation a closed case). 12,000 rpm is not the way to solve this problem, it's way too fast for brass at any tool diameter. And your machine doesn't go slower than 6,000 rpm, so 6,000 rpm is one of those rigid facts that can't be changed. So is your 2 mms diameter tool, since you want delicate lines that don't allow a larger diameter tool. Thus the only variable is the forward working speed. A better result at a second pass was already a strong hint; in fact you "clean up" the rough result of the first pass with some additional cuts, many of which land in between the spots of the previous cuts.

I have a suspicion that the cutting tool you used is not as sharp as it should be. You should be able to literally split hairs on the edges, copper based alloys only look (and are) nicely finished with super sharp tools. After being sure of decent tool sharpness, devide the forward speed by at least half, a quarter to 10% is even better. Than look at the result again; you'll be surprised. If the results are still not improved to a high degree, your specific sort of brass allot needs a slower rpm, meaning a change over to a different machine capable of lower rpm.

Baking the acrylic may discolour the brass, leaving you with additional polishing work that may also attack the freshly made acrylic fill. Certain lacquer formulas do exist that do not shrink so much while drying and curing, yet flow as a liquid when poured fresh out of their sealed container. You could fill the machined recesses with that; the lacquer will spread across the entire recess and creep into every tight edge all by itself, you may even top up the recess with additional fluid when the level is a bit too low after the initial spreading-out of the first fill. A long fine art brush with a fine point is ideal to let the fluid flow into the recess, so is a fine pipette. You may want to visit model maker and restorer forums to pick up some more tips and hints on this subject.

Lots of success!

gerhard

Thanks for the feedback guys.

Gerhard, I believe you have it with the tool sharpness. I did try faster feeds with more shallower cuts but ended up breaking the bits I had. I will be purchasing some new smaller cutters given I have only 1 2mm left, so my next job should have no problems with a lack of sharp cutters.

/M

Hi Mike,

what router are you using, and what software for the file designing & G-code creation?

Have you considered using 2-pack (2K catalysed) urethane paint, injected into the recesses with a syringe?

Yes, urethane is an excellent idea. Tough yet just flexible enough to shrink and expand together with the brass in cold and heat. Modern urethane formulas are relatively UV-resistant and keep their gloss and colour for a long time.
And a syringe is even better than a pipette, provided the drugstore will let you have the needle as well. Here in Holland, scenarios like drug addicts throwing the things in an alleys after use, endangering playing children, are thought up more and more, so in a Dutch drugstore or pharmacy i would have to explain the use before they'd let me have a complete syringe. Things are toughening up here.

As for the engraving pic, the machine pictured also turns at a few thousand revs, but the working speed is very slow and precise, which is visible in the clean smooth edges of the finished typefaces.
The flexible nozzle blows away cuttings with compressed air. Cuttings can also be sucked away (especially with the machining of plastics like acrylates), but such nozzles are of slightly different design.
Air cooling doesn't do very much for the bit itself, a minor part of its cutting edge heat is dissipated by its shaft and the major part is drawn away by the brass, which has good heat conductivity (one of the reasons for brass allowing dry machining).

The engraved plate in the pic could be brass, but am i wrong or could this also be a hue of bronze? The source where i got the pic from, didn't specify this. Whatever, with bronze being harder than brass, working speeds would be even lower. In both metals the same edge smoothness can be achieved, as long as the cutting speed and tool sharpness is right.

regards

gerhard

Hello Gerhard,

Thank you for a very useful and informative series of posts. I very much appreciate the information you have shared.

Hi Stewey,

Artcam to draw and create the gcode.
I didn't think of 2 pac, but I tend to avoid these as you have to mix part A and B and I always find it hard to mix the ratios. Maybe some digital scales would help there.

The pics that Gerhard shows are what I suspect I would aspire to achieve. But what i don't understand is, commercial operations must blanket cover the sign and then buff the paint off of the top, otherwise it would take ages to get it into each groove - especially the bee picture which is fine and complex.

It's good to understand that coolant like kero isn't required as this is messy to use on a flat bed machine like mine, with no specific drainage.

It's funny, i see a lot of brass signs about and i now take a moment and study them looking for information on the machining and how they may have been routed. A bit more practice is what i need i think.

thanks again for replies
Michael

Hi MIke,
you can buy graduated mixing containers from decent paint stores & Mitre-10 etc.

Alternately, I pour the needed amount of part A into a can, then dip a hacksaw or paddlepop stick in, withdraw it, measure the depth, divide by three, and add that to it, & make a pencil mark there.
Then put the gauge back in, & add in the catalyst to that mark.

That gives you a fair 3 + 1 ratio.

Work in mm and you can do it in your head- eg if you have 15mm of depth, put a mark at 15 + 5 = the 20mm spot & fill up to there.

Hi all,

yes, that sounds like a good way to get the mix right by measuring volumes. Weight is also a common method used for mixing ratios. For that you need to know the specific weights of both fluids (e.g. grams per cubic centimetre). With many professional formulas, factory filled net product weight is mentioned beside the fill volume. If not, any decent factory customer service should be able to cough up such information, since mixing by weight ratio is common practice in many professional circles.

If you have to calculate the specific weight all by yourself, substract tare (empty) package weight from total full package weight to obtain net product weight, to be divided by the product volume in the package. Write the numbers for base fluid and curing/hardening fluid down somewhere and after that initial investment, it's a cinch every next time you have to mis those fluids.

Take a digital balance (capable of .1 gram steps) and any container of which you have established the tare weight beforehand. Then just watch the display while pouring both fluids, add one simple mixing stick as a sole gadget, and presto, you're done.

As for filling routed recesses, filling each part separately is less troublesome than pouring paint over the entire surface, raking the excess off somehow and buffing off any remains after hardening.
Colour filling is done by hand as a rule. Technically it boils down to topping up each recees, just like topping up a glass to its rim. Dosage machines -to assist with large series productions- do exist, but are still operated by hand. These machines look like an electronics box with a pigment container attached on top. There are adjusting knobs at the front and a small hose with electric wires running parallel. Attached to this hose and wires is a grip with a needlelike feeding device and an adjustable sensor. The needle is placed onto the bottom of a recess, a trigger switch is squeezed and a little pump feeds the fluid pigment through the hose to the needle tip. The capillary properties of the fluid are counted upon for it to quickly creep and spread across the entire recess surface and into each tight corner. The recess is then just topped up to its rim, the right moment of which is sensed by the sensor, which switches off the pump. Topping up each recess to its rim means having topped all recesses to the same level. This means that the lowering of the fluid level during drying and hardening will also be the same in every recess, which guarantees an all over even end result of the cured colour fills, for the entire workpiece.

It is possible to after-treat the brass, both chemically (to dissolve any spilled bits of pigment) and mechanically (slight buffing with soft but dense materials at the slighest working pressure, since the "digging out" of colour fill from the brass recesses must be avoided), but when the job was neatly carried out and the brass already was nicely polished, after treatment is seldom necessary.

Total covering with masks is not done as a rule, because the results can be unpredictable. If a sticky foil is used, the router bit will cut both the foil and the brass underneath. This may proceed smoothly, but on the other hand the slighest bit of fraying or curling up of the foil or letting loose due to heat, means a crevice for the pigment to creep in between. Such "creep blooms" can look like nasty spills and may take a lot of extra work to clean away.
Also, chemical stuff like wax or resin coatings, would take longer to get rid of afterwards, than the separate hand filling of all the recesses would take.

The bee example was engraved in a seal ring (to wear on a finger), so not only the line width but also the whole embellishment is small scale. Liquid fill is impossible for such scale, for this a paste-like pigment is used. This is smeared into the recesses and the excess is literally raked off with a tiny rubber rake or a non-textile type of smooth cloth. Any remaining stains or hues are buffed away after hardening. This remaining soiling is often minimal, a simple rub upside down across a table linen would already do to get rid of it, so there's nothing drastic about these small paste-filled engravings.

regards

gerhard