Acceleration, velocity, distance, and other fun Plasma calculations

[Simko]

I am looking for any information on recommended accelerations for plasma cutting and sources for how to calculate what a table might be capable of based on gantry weight, motor size, etc.

To calculate it perfectly would require some crazy engineering calculations, but I am hoping that there are some websites, calculators, XLS files or white papers on making an estimated calculation of what a combination of parts is capable of.

I have read in a few places that at least 35 in/sec/sec is recommended. Does this sound about right?

So if I am thinking of this correctly, in one second, the torch will be going 35 in/sec or 2100 IPM. Breaking that down further, in 1/2 second, the torch will only be moving 17.5 in/sec or 1050 IPM, still plenty fast. 1/4 sec is 8.75 in/sec or 525 IPM, 1/8 sec is 4.375 in/sec or 262 IPM, etc...

SOOOO... If your desired cut speed for the material you are cutting is 300 IPM (5 in/sec) and your acceleration is 35 in/sec/sec it will take 0.1429 secs to accelerate to 300 IPM.

Time = (Vfinal - Vinitial) / Accel
Time = (5 in/sec - 0 in/sec) / 35 in/sec^2
Time = .1429 secs

The distance traveled in that time is

Distance = (Vinitial x time) + 1/2(Accel x time^2)
Distance = (0 x .1429) + 1/2(35 x .1429^2)
Distance = 0 + 1/2 (.7147)
Distance = .3574 inches

So is this telling me that at 35 in/sec/sec, the first .3574" of the cut will not be performed at the required speed (300 IPM) and will be poor quality (divot?). If this is true, then why don't you see poor quality every time the table has to change directions (corners, sharp radii, acute angles, etc)?

Here is an example of where I am right now. If I build my gantry the way that I originally planned using 8016 extrusion, it will weigh in at about 175lbs with everything on it. I am planning on running NEMA 34 motors with 3:1 reduction. What acceleration am I capable of? I'd rather not build it first only to find out that it is not fast enough. If I use 8080 extrusion instead, I can shave 24# off the weight of my gantry, what does that do for me? If I can cut up to 400IPM with adequate acceleration (35 in/sec/sec?) then I would rather have the larger gantry as I would like to make this a dual purpose table.

[acourtjester]

Maybe you should look at Hypertherm's manual to find the cut speeds. 400 ipm (tin foil??)

[Simko]

Maybe you should look at Hypertherm's manual to find the cut speeds. 400 ipm (tin foil??)

Maybe you should read the manuals...

Page 3-35 in the Powermax 65/85 manual shows speeds of 350 to 500 IPM to cut 16 to 26 ga mild steel using 45A shielded consumables....

[acourtjester]

Productions setting good luck

[Simko]

Productions setting good luck

350 IPM for 16ga mild is under the 'best quality' column, not the 'production' column. I didn't think that designing for 50 IPM over the requirement was out of the question.

I am the new guy here, if I am way off base, I'd rather have someone explain what speeds I should be designing for rather than just responding with a crappy remark basically saying I'm an idiot for wanting my table to run 400 IPM with good acceleration.

[Simko]

Let me ask the question another way...

If I have a table with:

- a 175# gantry
- NEMA 34 motors on the X, Y and A axis
- 3:1 belt reduction on X, Y, and A
- profile linear bearings for motion
- CANDCNC electronics package for up to 200# gantry with Ethercut and DTHC IV
- Hypertherm PM 85

What is the thinnest material I will be able to adequately cut with good quality and detail?

[acourtjester]

First of all I never intended to say you were an idiot, you asked about cut speeds and I referred you to the Hypertherm manual for that info.
My comments were made in Jest not to insult my post are to help others no to pick a fight.
There have been many post about speed and cutting quality here dealing with motors, drive ratios, weight, and other factors. One of them boils down to cost if you have a high dollar production table you will attain high speed and good cut quality. Then there is a lower cost table that will not reach the level of the production table but will still give usable speed and cut quality. You have done the research to gain the knowledge and did not like my simple answer. I am a realist after all the work is done what you end up with will answers to the questions.
Only you can decide when your happy.

[jimcolt]

Plasma, as we all know is a speed dependent process. It also requires a wide range of speed......today you may be cutting 1" steel at 13 ipm, tomorrow 26 gauge at 450 ipm. Cutting thick material at low speed is relatively easy. What separates well engineered cnc plasma machines from the rest....is the ability to have high acceleration that can get you to that maximum speed in the shortest distance without losing track of the following error (which is a measurement between the machines intended path and the actual path). Most machines can go beyond their design limits without breaking....but will produce wavy cuts with corner overshoot and bad tolerances...along with out of round holes.

There is no single number for ideal acceleration. On high end industrial ($100k plus) cnc's there are acceleration settings that vary for speed ranges that can compensate for weight of the moving parts, backlash, flex, etc. I run one of my home shop machines at .3 G for most cutting applications to about 250 ipm, above that I need to back off to a lower number.

High end cnc's also require tuning of the proportional, integral and derivative gains....in order to fine tune the mass to the drive capability. This takes time, and to do it correctly takes some sophisticated equipment that can track the real time movement right at the torch mount and provide feedback data.

Jim Colt Hypertherm

[Capstone]

To answer you question first...

If I have a table with:
- a 175# gantry
- NEMA 34 motors on the X, Y and A axis
- 3:1 belt reduction on X, Y, and A
- profile linear bearings for motion
- CANDCNC electronics package for up to 200# gantry with Ethercut and DTHC IV
- Hypertherm PM 85
What is the thinnest material I will be able to adequately cut with good quality and detail?

Depends a lot on the rigidity/size of the table and the "Z" setup. The described set-up above doesn't state the "Z" portion so I can't speak for that, but I use the v-groove/roller bearings of Precision Plasma, instead of much more expensive linear guides. I use CandCNC and have a HT PM45. I also have a 175# gantry, and that's the speed killer here. As Jim states, Moving that thing around quickly then making it change directions at those speeds creates an enormous amount of force and cut quality WILL suffer. My table is home-built and small, so the gantry weight is the single biggest problem. I can't cut anything cleanly above 120ipm. I must use 30amp consumables on my HT PM45 for 14ga and 16ga, but I've never even attempted 18ga or thinner. I cut at 95ipm on 14ga and still get a lot of dross, but at least I don't have problems with "shudder/wiggle" in my cuts as a byproduct of every sharp turn in my pieces, which is a much bigger problem in cut quality.

Are you building this table to produce a high volume thin metal product that has astronomical profit margin not yet seen already on the shelves of Wal-Mart?

The possible and the "good enough" are separated solely by the amount of money you are willing to spend. Having a table cutting 400ipm and still having good cut quality goes beyond the capabilities of my table more than 3-fold, but the time, money and effort required on my end to improve upon my table simply can't be justified for a whole host of reasons. I only point that out because my table in all of its limitations, still amazes the average consumer with it's precision and at the end of the day, the real value is in the operator and his skills to design/execute what his customer wants and still make money. Anytime you're talking about a home-built plasma table capable of good cuts on thin metals at recommended cut speeds, there is a serious drop off in the return on investment beyond 300ipm. Large scale Laser and waterjet operations out of China have gobbled up the market of cutting 18ga and thinner metal IMO for anything that's kitchy as far as seasonal decorations etc. You will have a hard time competing with any product that fits in that box unless you've already got some product that no one else offers and you know that you must make dozens at a time. There's probably 30-50% of additional cost to reach that next level 10% ceiling of performance to attain 500ipm.

If you have deep pockets and have set a goal to reach these speeds just to say "I can cut at 500ipm" then it's your money and good luck, but I don't think anyone on Plasma Spider, not one, will tell you it's necessarily good money spent based on the thousands of cutting hours all of the members have already put behind them and the quality that their customers still pay good money to produce.

[Simko]

Thank you for the great responses...

I am just building this table for hobby use and maybe to make some parts on the side for people for extra cash.

I don't have deep pockets, but I have been known to build/buy/design above the level that I 'think' I need to cover any future requirements

I have learned over the years that when you get into a new hobby, you often make a lot of the same mistakes as other people. I am trying to avoid making the same mistakes as others by doing some research. Between this site and CNCZone, I compiled what I 'thought' to be slightly above average requirements for my table. Maybe what I was actually reading was people who have no idea what they are talking about and have 'big d!@k syndrome' of "My table goes 3000 IPM and can cut 30 ga with no dross" or "Your table has to accelerate at 100 in/sec^2 or you're a loser."

Here is where I am with regards to weight (174.8#). I calculated all of the numbers except for the ones marked with an *. Those were given to me by either Velox, CandCNC, or CNC Router Parts. The weight for the torch and cable seems high to me, but that is what I was told. The estimated gantry carriage weight also seems high to me. I could instantly shave 24# off and be at 150.8# total weight by going with 8080 extrusion instead of the 8016 extrusion. If the carriage and torch are estimated too high, that will bring my total weight down even further.

- 8016 crossmember - 79.75" long - 52.8 lbs
- Side plates - left and right alum - 20.3 lbs
- 8080 riser stiffeners - left and right - 5.7 lbs
- NEMA 34 PRO R&P - left and right with belt reduction and motors - 25.2 lbs *
- Gear rack - 79.75" - 9.8 lbs
- NSK LS25 linear rail with two trucks - 16 lbs
- Gantry carriage with IGUS track - 12 lbs estimated *
- Torch and cable - 15 lbs *
- Velox z-axis - 15 lbs *
- Velox torch mount/ mag break-away - 3 lbs *

[Simko]

I run one of my home shop machines at .3 G for most cutting applications to about 250 ipm, above that I need to back off to a lower number.

Jim Colt Hypertherm

Would this be the proper conversion from g-force to acceleration in in/sec^2?

1 g = 9.80665 m/sec^2
.3 g = 2.94 m / sec^2
.3g = 9.646 f / sec^2
.3g = 115.75 in / sec^2

If so, that would explain why the table can produce clean cuts at the corners because it only takes a distance of .07499" (just over 1/16") to reach a cutting speed of 250 IPM.

[exapprentice]

Hi Simko

I designed and built my own 10' x 5' table with the same intentions, unfortunately as already said before going the DIY route is one of personal choice.

Part of that choice is how much you want to spend, I believe that there is a tipping point regarding cost and that is, once you reach a value that is close to buying a ready made table and all that comes with it, why spend more on something obviously your not sure about.

My intention was to build the best I could for as little as I could and that meant many hours of my time (hobby time) working it all out, shaving weight here and there, machining my own parts, doing work arounds to save costs and building my own Z axis.

I have my own engineering shop that meant I could make parts in my own time and save the costs of having bespoke parts made for me, there are lots of people on this site who do not have the same facilities but they have still gone on and built their machine and are pleased with the results, thats what I think DIY is all about.

At the moment I can cut 20swg CR4 and to be honest I am more than pleased with the results watching something you made moving at 8100mm/min is awsome but it does hi-light all the inherent weaknesses of my table, I now know where and what needs improving, so that "To Do" list is always live.

At some point you will have to put that stake in the sand and start your own build (if thats your chosen route) otherwise you can just keep on designing and re-designing but not actually know if it works (OK if your a machine builder who has to be sure, but even then there are no guarantees)

Every DIY build is different, thats what its all about and I am sure the majority of people on this site have made mistakes building their tables or wished they had done something differently (always have a plan 'B')(or 'C')

Good luck with your build, looking forward to your future posts on your progress, it will be interesting
and the build in my opinion is the easy bit (based on my total in-experience so far and that learning curve just keeps getting steeper)

Best regards to all

exapprentice

[Black Forest]

I didn't see listed what you will be using for driving the gantry. Meaning, rack and pinion, belt driven, ballscrews, etc.. Also you listed NEMA 34 motors but not what type, stepper or servo?

Lots to learn in this CNC plasma journey, Grasshopper! That is what makes it interesting. I ask these questions so I can learn also.

[Simko]

At some point you will have to put that stake in the sand and start your own build (if thats your chosen route) otherwise you can just keep on designing and re-designing but not actually know if it works (OK if your a machine builder who has to be sure, but even then there are no guarantees)

Best regards to all

exapprentice

Thanks for the insightful post... I agree with you that at some point you just have to do it and make modifications as necessary as you go. I am just trying to make sure that my 'stake in the sand' is as close as possible to where I actually want to end up.

I didn't see listed what you will be using for driving the gantry. Meaning, rack and pinion, belt driven, ballscrews, etc.. Also you listed NEMA 34 motors but not what type, stepper or servo?

The X/Y/A axis motion is going to be with done with profile linear bearings (NSK), rack and pinions, NEMA 34 steppers (980 oz-in) with 3:1 belt reduction. The z-axis is going to be a Velox z-axis direct coupled with a NEMA 23 stepper (620 oz-in).

I looked into servos, but I don't feel that I can justify their additional cost at this point.

[tcaudle]

Now for some physics....
A= F/M
Acceleration = Force (Ft-lbs)/Mass (weight/32)

Force is the LINEAR force provided by the motors moving the gantry. You have to convert the oz-in of rotary torque by dividing the final drive element pinion gear by its radius in inches. You have to calcualte the torque gain from belt reduction then you need to convert that to ft-lbs of force to be in the same units as the MASS BTW you cannot use the holding torque of the motors . Its a lot less at 50% RPM (roughly 50% at that point) but it drops off rapidly as yu go up in RPM. As you cansee the faster you move the less acceleration it will have simple from the motor RPM (unless you are usng servos).

Once you have the ft/sec/sec of acceleration you convert that to in/sec/sec (or G's )

So lets say you have two 980 oz-in steppers with a 3:1 belt reduction and 1" diameter pinion gear. Top speed is about 700 IPM so cutting speeds should be kept around 350 or less. BTW you cannot use the holding torque of the motors . Its a lot less at 50% RPM (roughly 50% at that point) but it drops off rapidly as you go up in RPM. At 350IPM you have roughly 490 oz-in per motor times two (motors) so 980 total . Belt reduction gives you 3X so you have 2940. Divide that by .5 inch pinion radius and you have 5880 oz-in of linear force ....not bad. or 30.625 ft-lbs
Now divide that by 5.648 (Mass in SLUGS) you have roughly 5.42 ft/sec accleration....still respectable at 65 ips/sec

Before you get too excited about the numbers there are a lot of losses in the mechanics so a PRACTICAL number is about 1/3 to 1/4 of the final acceleration. Now you are down to 21.68 ips.sec in REAL numbers. Which says you can't expect sharp corners and turns at anything over about 200 IPM. with taht combination. Bigger motors are not alwyas the answer because the upper RPM range of bigger motors is less so their torque curves say they will acutually have LESS torque at 350 RPM

Now start playing with the numbers and see what weight it takes to get you closer to the 35 to 50 IPS


As youy can see the weight direclty effects the acceleration at a given torque level.
Heavier masses also bring greater inertia (Newtons Law ) and it creates large counter forces that show up as something that has to be assorbed by the table structure. So even if you find you can move and accelerate a 175 lb gantry at 350 IPM and have acceleration up to 35 IPS/sec (actually at 350 you need closer to 50 IPS/sec for sharp turns) you have to have a table structure that will take that and not flex/vibrate too much. What causes a lot of problems is your material (which has MASS too) is laying on slats that are spingy and when the table flexes from the ineritalof the sudden moves the slats flex and the inertial of the material causes it to move at a different rateas the table structure and you get a nice little wavy scallop cut

Every pound you can take off the ganry the more acceleration you can get , and the less inertia and the less vibration on the table.

One of the things you missed on your numbers is while the time to move to full speed is correct you are already moving at 50% of full speed in half the distance and divots don't happen when you are moving, just sitting still. A divot takes finite time to form where the flame gets straved for material and starts cutting the sides and back.

[Simko]

Now that is what I am talking about!

Thanks Tom for taking the time to explain that. Before visiting the site tonight, I decided to order the 8080 extrusion and shave 24# off of my gantry. Worst case, I will be at 150# but I think after it is all said and done, I will most likely be around 130-135#. It won't be as solid for routing, but routing is not its primary purpose. Anyone have a need for an 8' long piece of 8016 extrusion (80mmx160mm)? LOL

I should also thank you for taking the time to talk to me the other day on the phone for 10-15 mins to discuss my options for electronics

My table is SOLID! There won't be any flexing and if the forces from the gantry are enough to move the entire table, I'll lag it to the floor.

Time to go play with some numbers.

Thanks!

[Simko]

What is the variable that says that 21 in/sec^2 is adequate for 200 IPM cutting, but 50 in/sec^2 is required for 350 IPM?

Is it the distance required to get up to full cutting speed and if so what is the largest target distance?

21 in/sec^2 at 200 IPM takes .2645" to get to full speed.

50 in/sec^2 at 350 IPM requires .340" to get to full speed.

35 in/sec^2 at 350 IPM requires .4861" to get to full speed

Based on these, it looks like the target would be somewhere around .25 to .33" or less to reach full cut speed to get quality cuts.

[jimcolt]

The target distance to accelerate to ideal cutting speed is 0" (mm). The longer the distance, the longer of an area with speed related effects on the cut. Low speed causes: dross, wider kerf, larger heat affected zone.

No machine is perfect, however it is important when your intended cut parts need best accuracy and detail to get acceleration rates optimized for your particular machine design. Light weight and rigid designs with no backlash are the keys to success. On most machines their will be a limit to acceleration rates that will be indicated by waviness, ripples, and excessive cut edge striations......so you generally tune your machine tighter (higher acceleration) until these features appear (on holes and exiting sharp corners) then back off on your gains or acceleration rates until the cuts are acceptable. If you cannot "tune out" the cut edge anomalies, then they could be caused by: undersized or inertia mismatching of the motors, flex in your machine mechanics, backlash, or binding in the motion (overtightened pinion (spur) gears are often a culprit).

Try to think of what the plasma is doing in comparison to a router bit or an end mill. It is melting and blowing away metal at a fixed rate......so when speed slows the plasma arc just burns more material, making a wider kerf and adding more heat to the material....which will cause dross and warpage. Ideally when a machine decelerates you could reduce the plasma cutting power (amperage) ...which can be done with limited success. (keep in mind that to reduce the cutting amperage mid cut with ultimate success....you would also need to change the nozzle orifice size and plasma gas flow in order to maintain ideal arc energy density).

Listen to what Tom Caudle says! He has the unique position of working with his customers....thousands of machines built with Tom's cnc drive systems with no two of them exactly alike! Different sizes, different weights, different drag numbers (from motion ways, cable tracks, etc.)....Tom probably has more practical experience with a wideer variety of machine designs than anyone in this industry...


Jim Colt Hypertherm