[3d_cristian]

Hello all, for the last few years I have been designing a CNC machine that has the following features:
- spindle is powerful enough to be used for steel milling using typical end mills size and speeds/feeds as per tool manufacturer recommendation.
- easy to build with no special instruments, however it can produce components within 4 thou (0.1mm) tolerance.
- the machine size and the work envelope can be easily customized.
- low weight, on wheels, in order to be easily moved to a different location or inside a shop.
- the Z axis containing the processing head can be exchanged in order to use the machine for multiple manufacturing processes: milling, laser cutting, carving, etc, etc
- it can accommodate any motion card, CAM software or operating system (Linux or Windows)
- ATC with 5-10 tools
-vibrations produced during machining will not have an impact on the machine performance, maintenance or lifecycle.

Recently I have finish the prototype and posted the frame CAD files on Grabcad https://grabcad.com/library/powerful...teel-milling-1
During the prototype build iterations I have tested and recorded the vibrations levels for milling aluminium and steel. I will follow up with more vibration measurements shortly.

[awerby]

Have you actually built this thing and tested it on steel? If so, could you post some actual photos of it, and some videos of it cutting steel? If not, how did you manage to quantify its accuracy?

[joeavaerage]

Hi,
seeing is believing. From personal experience cutting steel or stainless is a WHOLE different experience than copper/brass/aluminum
and I've yet to see a 'router' style CNC machine do a credible job.

Craig

[3d_cristian]

Sorry for late response. I am working on video and vibration analysis in the same time. Formatting the results will take some time.

[3d_cristian]

Hi Andrew and joeaverage,
I have started to add videos to my youtube channel. The video and audio quality will get better on the next posts.
There are few comments that I would like to add:
- There is a distinct difference in sound between steel milling (at 3000 or 4000 RPM) and wood or granite carving (24000RPM). The spindle used for steel milling has an instantaneous peak torque of 30Nm (over 22FtLb).
- You can see from the are videos that are done on the same machine which I have built few years ago. Based on my design I can exchange the spindles based on the manufacturing process ie: steel milling vs wood carving.
Let me know if you are interested to see more capabilities of the machine.
As promised, I the next few days I am planning to add more videos in the same youtube channel, showing the precision of the machined parts and vibration data

https://youtube.com/channel/UCCArc-Er8QKMQqW0XLsU7Jg

[Andreiir]

Hi.
When i saw this tittle i was thinking lets see a nice precise and stiff , and powerful construction .
But when i saw design and the precision requirement , i was thinking WTF is precision now , or power means to people ,
i really dont get that words POWERFUL and PRECISE .
In ur world precise is 0.1 mm ?
And powerfull is some carbide tool scratching a steel sheet ?

[Zorbit]

The wobbly stacks of aluminium extrusion don't look suitable for a steel cutting machine to me.

[3d_cristian]

Hi Andreiir and Zorbit,
I have friends that invested more than $1M to buy milling machines. They are very scared that some of their jobs will be taken away from guys that invested only $10K on milling machines that are only 300Lbs and can be easily moved from shop to shop.
I am a mechanical engineer with many years of experience. Tolerances of 0.1mm or tighter are required in some industries that you can count on your fingers. Usually they are big corporations that will not work will small/medium shops.
Vibration and thermal expansion of metals are a fact of life. However, over the years, people found ways to overcome these problems and to make railroads, bridges and skyscrapers. I am not taking credit, I am just just using it as an example.
The information that I am posting are just final tests of my design with no information about how I have solve some of the issues. This way everybody can decide if they like or do not like the results based on their particular application.

[awerby]

There's a lot of distance between a million dollars and ten thousand. There will always be small shops with less invested capital that will be able to take on jobs that aren't economical for the bigger players. But if the job involves machining things out of steel rapidly and to a high degree of precision, I don't think your router design will cause them to lose any sleep. While boxing in the Z axis like you've done, and using side walls instead of perching the gantry on high stilts does add a certain amount of extra rigidity compared to a typical router it's not nearly enough to make this into a real milling machine. It might be able to handle aluminum to some extent, although the videos you've posted so far aren't very convincing even of that.

[Zorbit]

As a machinist I often worked to better than 0.01mm. I can't recall ever having a job with 0.1mm tolerance for milling. I'm sure many jobs would have been fine with 0.1mm tolerance, but I don't remember ever seeing it specified. 0.05 was quite common.

[joeavaerage]

Hi,
my current build mill has cast iron beds, 115kg each, and a 300kg SG cast iron frame. My FEA analysis suggests a very credible 220N/um stiffness in X,Y and Z directions.
Just the cast iron parts will when finished cost over $10,00NZD. I designed it to do a nice job in steel but could still only be called light duty, genuine production CNC
machines are 1000N/um and better.

Quite frankly if I thought a gantry style router were stiff enough to mill steel I would have built one, but I needed a bed mill to achieve the rigidity I wanted.

Craig

[machinedude]

if you think about the cost of a build to end results this machine is doing good for the investment cost and i would keep at it to make improvements. one thing i did notice was the surface finish was not very good. the machine looks to be cutting fine for what it is but the finish results are showing some room for improvement.

where i could see something this light and mobile being useful would be in repair work in the field. so being able to hold +/- .005 out in the field doing repair work on site would be something i could see being useful. So i think something light duty and mobile could have it's place.

one thing i noticed with the t-slot extrusion is it tends to not deal with twisting type forces very well. if you fill the voids and unused t-slot with E.G. it will help quite a bit in that area and not add to much weight. the E.G. will also dampen a lot of vibration. might help with surface finish as well.

[machinedude]

As a machinist I often worked to better than 0.01mm. I can't recall ever having a job with 0.1mm tolerance for milling. I'm sure many jobs would have been fine with 0.1mm tolerance, but I don't remember ever seeing it specified. 0.05 was quite common.

i think it depends on the type of work your doing. the bigger the parts the the more the tolerances start opening up as a result. most parts might have 1 or 2 features that have to be close but the rest of the part has large allowances for tolerance.

the thing about machining is there is such a wide range of applications for the process that you end up with a wide range of tolerances as a result. i've been doing it for around 30 years so i have seen my fair share of the industry.

i have actually seen tolerances as much as +/- .100 imperial on large weldments now granted they were 4 foot in diameter and about 3,500 lbs or better but things can get pretty generous on some stuff. not everything but most of it anyways.

[jackjr-123]

Hi,
my current build mill has cast iron beds, 115kg each, and a 300kg SG cast iron frame. My FEA analysis suggests a very credible 220N/um stiffness in X,Y and Z directions.
Just the cast iron parts will when finished cost over $10,00NZD. I designed it to do a nice job in steel but could still only be called light duty, genuine production CNC
machines are 1000N/um and better.

Quite frankly if I thought a gantry style router were stiff enough to mill steel I would have built one, but I needed a bed mill to achieve the rigidity I wanted.

Craig

You might want to redo your FEA analysis because these numbers aren't credible at all. It's very easy to get inflated numbers when you don't know what you're doing.
For reference a basic Bridgeport is between 10-15N/µm in static stiffness, for a weight about 1-2 t.

[joeavaerage]

Hi,
thanks for that input, I will double check the data units.

I dubious about 10-15N/um, 10-15kg force per um I would believe however. I have been caught in that confusion before and as a consequence am very particular about
consistent units.

Remember also that a Bridgeport is much larger, approx double the dimensions of my machine. Correlating stiffness and mass only makes sense with machines
of the same size.

Craig

[jackjr-123]

Hi,
thanks for that input, I will double check the data units.

I dubious about 10-15N/um, 10-15kg force per um I would believe however. I have been caught in that confusion before and as a consequence am very particular about
consistent units.

Remember also that a Bridgeport is much larger, approx double the dimensions of my machine. Correlating stiffness and mass only makes sense with machines
of the same size.

Craig

My units are correct. I really meant 10-15N/µm.
For example here's a study about VMC stiffness: https://journals.sagepub.com/doi/ful...87814016656533
Results: about 25N/µm total static stiffness for a typical VMC (50N/µm, 100N/µm, 100N/µm for X/Y/Z).

[3d_cristian]

Hi guys, I was able to measure vibrations while milling steel. The video is called "Vibration monitoring for steel milling" and is located at my youtube channel https://www.youtube.com/channel/UCCA...XLsU7Jg/videos
There are another 2 videos that show the calibration of the accelerometer.
Also, the machine design in Grabcad has reached 160 downloads. The files are located at https://grabcad.com/library/powerful...teel-milling-1 and https://grabcad.com/library/powerful...teel-milling-2

[3d_cristian]

My units are correct. I really meant 10-15N/µm.
For example here's a study about VMC stiffness: https://journals.sagepub.com/doi/ful...87814016656533
Results: about 25N/µm total static stiffness for a typical VMC (50N/µm, 100N/µm, 100N/µm for X/Y/Z).

I would love to add comments to your calculations however I cannot reveal design data. Please use testing data instead.

[3d_cristian]

if you think about the cost of a build to end results this machine is doing good for the investment cost and i would keep at it to make improvements. one thing i did notice was the surface finish was not very good. the machine looks to be cutting fine for what it is but the finish results are showing some room for improvement.

where i could see something this light and mobile being useful would be in repair work in the field. so being able to hold +/- .005 out in the field doing repair work on site would be something i could see being useful. So i think something light duty and mobile could have it's place.

one thing i noticed with the t-slot extrusion is it tends to not deal with twisting type forces very well. if you fill the voids and unused t-slot with E.G. it will help quite a bit in that area and not add to much weight. the E.G. will also dampen a lot of vibration. might help with surface finish as well.

Hi "machinedude", you are correct on:

1.)The surface finish was not very good indeed. I do not remember if I used a finishing pass or the tool left "scratches" because I did not restrained the X axis very well. I have revised the design in meantime and new I can machine mild steel.

2.) The t-slot are not great at twisting force. However, there are no twisting force present on my machine because of "closed frame" design. Also the design is very flexible and I can easily add another extrusion for support if I can see it required. The one thing about my design is that I can add unlimited number of extrusions to strengthened it before it comes too expensive or too heavy.
The epoxy fill was a backup concept from the beginning and I can still do it while the machine is fully assembled. There is a lot of empty volume on the extrusions and the epoxy is not cheap. Now that I have a working steel milling machine I am just not sure is worth the extra money.

Cristian

[3d_cristian]

The wobbly stacks of aluminium extrusion don't look suitable for a steel cutting machine to me.

"The proof is in the pudding" Is it suitable now? https://www.youtube.com/channel/UCCA...XLsU7Jg/videos