Sign in to follow this  
Followers 0
Donovanr

High speed motion project

23 posts in this topic

I have a high speed flying shear project i am looking at with 2 or 3 axis synchronized at one time. Using the Motion Analyzer software tool you can calculate the motion profiles etc What tools can i use to ensure a Controllogix L6? processor will be able to process the commands to the K6000 servo drives via serco's in the cycle times required? I have been told that this application should be done using a motion controller and not a PLC! When do you decide to take the next step?

Share this post


Link to post
Share on other sites
Just to clarify, the L61 will not be doing the actual motion control. The M02AS is a motion controller. Short answer to your question, the L61 "cycle time" will not affect the time it takes to process commands to the K6000 Long answer, there is a whole lot more information needed to provide an answer. There are also many precanned controllers out there for flying shear applications. Hopefully Peter will come along and chime in, but it would be worth contacting him at Delta Motion about your concerns. His company developed the motion controller for AB and he also has an alternative product which may be of interest to you.. Also while I am unsure of who or how he is employed, Gerry who is also located in Auckland is very knowledgeable about the AB motion controllers. If you would prefer to talk to someone in person and go over your application, it would be well worth trying to contact him and see if he would consult you on your application

Share this post


Link to post
Share on other sites
If would be nice to know exactly what Donovanr is trying to do. Most shear applications require synchronizing 1 or 2 axes to a line encoder. Chances are have program already done that is similar to what Donovanr is going to do. I can program a controller to show how we can do Donovanr's application. The RMC75E is a good choice if you are synchronizing two axes to a line speed encoder. It has a 0.5 millisecond scan loop and the user program runs synchronously with the control loop. The advantage is that all the control is done within the controller so there are no delays transfering control information from the controller to the PLC and then back to the controller. Having everything self contained minimize the phase delays between reading the line encoder and updating the servos. We have flying shear programs done already. Here is a plot of a real flying shear in action. The traverse axis moves a 3000 pound shear. ftp://ftp.deltacompsys.com/public/Picture...ing%20shear.png The data is taken at 0.5 millisecond increments and shows one cycle. There is a color coded legend in the lower left. Red: traverse actual position Cyan: traverse target position Blue: traverse actual velocity Magenta: traverse target velocity These two lines start a 2 inches and then ramp up to match the line speed at where the vertical cursor is Black: shear actual position Orange: shear target position The shear comes down once the traverse synchronize to the line. Green: the control output to the tarverse servo valve. I am not clear about the other lines because I can't read the legend and this was taken 3 years ago. The RMC75 familly http://www.deltamotion.com/products/motion/rmc70/index.php Specifically http://www.deltamotion.com/products/motion/rmc70/cpu75e.php The RMC75e communicates to Rockwell PLC using Ethernet/IP. The RMC75e shows up as a SLC5/05 on on RSLinx. So the status and control data 'automagically' get transfered. Parameters are usually downloaded using MSG blocks. BUT, the RMC75e has a PowerPC and 64 MB of ram and is programmable using a subset of structured text. The point is that the RMC75e can run alone if necessary with just a HMI to show what it is doing. We have a RMC75e on-line so you can play with one over the internet. We have customer that have pretty much done their projects over the interent. There are always some tweaks that must be made on the real system. We provide on-line internet demos and training. We have a distributor in Hamilton, NZ. http://www.ecs.net.nz/m14.htm Older products are shown. Tech support is excellent if you have a phone and even better if you have internet access. We also have our own forum at foruim.deltamotion.com. Not much happens there. I like to keep it that way. Anyone with experience with motion control shouldn't have any problems with getting the controller to go. What ever you do, don't try to write the code yourself. It will cost you more time than what you save in hardware and a PLC isn't fast enough.

Share this post


Link to post
Share on other sites
Hi Peter an TW Thanks for your replies, Sorry for the brief explanation... It is a Rotary Shear application very similar to this link. http://www.galilmc.com/learning/motioncode...loaf_slicer.swf Except we have a possible 2nd X axis feeding while the 1st X axis is finishing. There will also be other axis involved but not as critical. I know of the Delta systems @ ECS, i will doing further investigation. Peter are these preloaded programs for Flying shear available I just wanted to know the limits of Controllogix Motion using sercos versus a dedicated motion controller eg. Delta , Siemens Simotion D, Elau Pacdrive.... Thanks

Share this post


Link to post
Share on other sites
The key is to minimize communication delays when the cycle times are short. I bet any of these could do a flying shear if the speeds are slow, but there are other things to consider. Ease of programming, diagnostics, tech support, documentation and I am here. We also have our own forum at forum.deltamotion.com but there isn't much activity there and that is the way I like it. That tells me that things are going well.

Share this post


Link to post
Share on other sites
Donovanr.... don't discount a pure Contollogix/Motion Card/Kinetix Drive solution. It's really a simple application with the external encoder wired to the Kinetix drive auxiliary input and a MAPC instruction cued up in the motion controller waiting to execute based on position.

Share this post


Link to post
Share on other sites
A video and example flying shear program http://www.deltamotion.com/peter/RMC75%20R...FlyingShear.zip I updated these files 090508 11:07 to correct my calculation of the slave ramp distance and to fix some comments and spelling. Donovanr, would need to adjust the distances and speeds to match his requirements. It looks like he may have other axes too but I can modify this program easily enough once I know exactly what the application needs to do. I would bet Donovanr's application needs to run much faster than my demo. If I made the system too fast it wouldn't be easy to see what was happening in the video. At the end I show how the real time trend works at 0.5 millisecond sample rate. The only time I find this mode to be handy is when it is used in conjunction with the trigger were the plot can be triggered and frozen like a logic analyzer where you can see 50% before and 50% after. Since I programmed my window to be 10 seconds wide the plot would store 5 seconds before and 5 seconds after. The ratio is programmable too. Let's see a PLC or even most motion controllers do that Also note that no cam table is required. All the calculations are done on-the-fly. Edited by Peter Nachtwey

Share this post


Link to post
Share on other sites
Hi Peter Thanks for your attached demo, well commentated!! I will attempt to watch this a few times to let it sink in... Is this demo created using a simulator in RMC Tools? How does the RMC controllers compare to an AB L61? eg scan time What drives and motors do you prefer to recommend to interface with RMC controllers? Cheers Donovan

Share this post


Link to post
Share on other sites
Again, I have know idea of how much mass you are trying to move so I can't recommend a drive but our customers have use the Kinetix 6000 servo drives you mention. We are non-denominational when it comes to drives,VFD and servos. Most customers use drives a VFDs but you should consider a true servo for the synchronizing actuator. Linear actuators like Exlars etc work well but they may be speed limited. A true servo system is best, not a VFD, for the synchronizing axis. The synchronizing axis must accelerate and decelerate quickly so you should consider spending a few bucks more there. You said you application is a fast one so I assume the accel and decel rate will be high.

Share this post


Link to post
Share on other sites
Donovanr...... her is a CompactLogix/Kinetic PLC program that demonstrates a flying cutter application as another option. This project ran 400mm + long Punch_Cutter.ACDparts ( Part length set on MMI ) at 30 meters/min.... roughly a part every 1/2 second. Make sure to open the graphic camming curve on the MAPC instruction. Good Luck hey !

Share this post


Link to post
Share on other sites
Donovanr, here is a link to another video that shows how we ensure our scan times are synchronous and deterministic. It goes "under the hood or bonnet" to show how we compile the code and count the time it takes each instruction to execute. http://www.deltamotion.com/peter/RMC75%20R...ls/ScanTime.mp4 Again, the scan time in the flying shear program is fixed at 0.5ms but you can that we use far less time.

Share this post


Link to post
Share on other sites
Thanks for your sample program WildPointer... Is your application a flying linear shear? Where the axis would synch with the material cut it then return to start position? Nice simple code

Share this post


Link to post
Share on other sites
Go to this link: http://samplecode.rockwellautomation.com/i...1&EndRow=75 There is a sample of a shear machine, called flyingshear sample Edited by widelto

Share this post


Link to post
Share on other sites
Hi Guys, There's been a bit here on the subject already but I've noticed there wasn't anyone who bit on the performance of the Controllogix L61. I can't comment on the performance of other systems but I've spent the last 5 years working with CLX motion. There are a few cycle times for motion in RSLogix5000 due to different tasks running on different processors/asic's: -Scan time of the PLC logic: This controls when a motion instruction can be initiated, usually with flying sheer this is not an issue as you issue a "pending" cam which you can initiate long before the cam needs to do anything and when the axis you're camming off hits a predefined point the cam takes over.. these are very repeatable and accurite and I have programmed flying sheer applications with them. -Course update rate for motion: This is the periodic (and high priority) task that calculates what the axis trajectories should be given the last 255 motion instructions issued against that axis. -Sercos update rate (make it the same as the course update): The rate at which trajectories and command are communicated out to the drive -Loop times in the drive: The actual closed loop calculations to maintain position, this is done in the drive and does not vary with the number of axes.. usually in the region of 128 micro seconds time performance? In terms of processor load the L61 can handle 31 flying sheer slave axes and one master, scanning all the code (including homing, abort code etc) and commanding the axes with 8ms course update (trajectory calculation). The speed you can send this out to your drives depends on what they are.. for a kinetix 6000 this should be half a milisecond per axis minimum but I suggest setting it to the same as the course update rate. Has it worked before? Yeah, flying sheer is a pretty simple motion task and should be no issues on an L61 unless you're getting down to really low (below tens of milliseconds) cycle times for cut initiation. Cheers, Mike

Share this post


Link to post
Share on other sites
Thanks for steering the train back on the tracks Mike... I have a bit more information about the application. Axis 1 is a rotary axis operating at 1500rpm, it takes 0.040secs per rev. 25kg blade and shaft which is a balanced load Inertia 0.4kgm^2 Diameter 800mm Axis 2 is a belt driven system that has to index 2.3mm in 0.013secs and dwell for 0.027secs 4x 70mm diameter pulley's with 0.00318 kgm^2 inertia Shaft inertia 0.00024 kgm^2 Mass of belts 1kg Gearbox ? Operation With Axis 1 operating at 1500rpm, we have 120deg per rev to index Axis 2 2.3mm and dewell 0.027secs for 240deg. There will be approx 140 indexes per cycle before it returns to home position. As we are dealing with 0.013secs to index i would like to ensure a L61 with K6000 is capable of this. Cheers

Share this post


Link to post
Share on other sites
You should have posted this information first. The speed isn't even that high but the acceleration is. I didn't see any number for the inertia of the motor. Have you done the math to figure out the g forces? I have. You are asking for a lot. I will bow out and the Rockwell guys should too. You need to really do some simulations. The real question is that can you make a system that can physically do this. It is easy to shuffle electrons around.

Share this post


Link to post
Share on other sites
That is exactly the same question i threw back at my customer... I have seen a similar machine that has had years of development, I think the motion control is handled using Elau Link

Share this post


Link to post
Share on other sites
Actually, we make machines that scan french fries and cut the defects out. The blades must be able to cut in 16 milliseconds. I took a lot of research to make knives that would withstand the forces. We had to have custom servo motors designed for us that had low inertia. We even had do design our own drives. This was very expensive. I doubt an off the shelf solution exists. I think your customer needs to be prepared to spend the big bucks to do what they want to do. I don't think you should stop the indexers and actually implement a double flying shear where there are two flying shears that cut every other slice.. Then you may have a chance. So, did your customer or you do the calculations to figure out the required acceleration? I calculate that the peal acceleration will be around 8g if s curves are used. I think you should walk away unless the customer is very serious. The project can be done but it will cost big $$$$$.

Share this post


Link to post
Share on other sites
Is this what you calculated for Axis 2? The blade (Axis 1) is a constaint speed, it can take approx 30 secs to ramp up to speed and maintain set speed. Using Rockwell's Motion Analyser i can size a MPLB... motor to carry out the profile with a 20:1 gearbox 8g sounds rather aggressive, have we missed something here?

Share this post


Link to post
Share on other sites
Glad we've got some numbers on the timing now. You're right to be careful with this. I won't comment on the physical side because there's much more familiar motion analyzer users than I but on the cam I think it's going to come down to the course update rate in the L61. There is a limitation with logix when the cam starts getting small relative to the course update, the CPU only puts out trajectories for the drive for times corresponding to the course update i.e. at the next course update the slave should be at position X. Also cams can only initiate at the course update rate. To make things absolutely certain you will have to: -Run small course update times, definitely below 6ms, more like the default of 2ms. This increases the resolution of the trajectory. -Make sure the master lock position (ordinate of the master axis where the cam starts doing something) happens on a course update event. The will start the cam deterministically. -Make sure your cam points all happen on course update events (or that there's a number of cam points between course updates, but this increases processor load.). This will make sure the trajectory is deterministic. Normally you don't have to worry about things like this if the timing is not so tight. Maybe you might also look at an indexing servo (Ultra3000 indexing?) as it would take the heat off the processor and possibly drop the cost/complexity.

Share this post


Link to post
Share on other sites
Ideally there should be an even number of coarse updates per move. 2 ms does not divide into 13ms evenly so how is one going to do a motion of 13 ms and a dwell of 27ms? The whole cycle can be approximated by a cam but the move will not be exactly 13 ms long.

Share this post


Link to post
Share on other sites
I guess I'm a little confused with the Coarse Updates/Trajectories. The Coarse Update Period is defined as: Selects the periodic rate at which the motion task executes to compute the servo commanded position, velocity, and accelerations to be sent to the 1756-M02AE or 1756-M08SE modules when executing motion instructions. So my take on it was that it is the rate at which we exchange data/instructions/trajectories between the PLC CPU and the Motion CPU. Once a "Persistent " or "Continuous " Cam is loaded into the motion CPU is it not now independent of the PLC CPU ? Does not the motion CPU update to the drive now at a very high rate much quicker than the Coarse Update Period ? I thought that all the trajectory information would be held internally on the motion CPU after executing the Cam instruction once.

Share this post


Link to post
Share on other sites
No.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0