Use cycle simulation time error

[pinguS]

Has anyone ever come across this weird time mis calculation...

In turning when I run a simulation with use cycle option for a turning operation, the time for the part says 2 min 30 second, but soon as I select semi / finish on with ISO turning rest material (bottom right of technology page) the simulation time jumps to 7 hours +.

I have checked the tool data, and have made the finish speeds and feeds the same as the roughing ones. SO what the hell could it be. The program from the G code seems right, but this timing thing is way out. Its not actually a problem at the end of the day, but just weird.

[pinguS]

Ok another thing I noticed is that when showing the simulation data, the feed rate showing for roughing is 900 (tool is set to 3000 revs and 0.3 mm/rev for both roughing and finish in same turning operation).

But when it runs the finish cut, it is showing data as 0.3 mm feed rate

WHY????

[Brakeman Bob]

The time taken in SolidVerify is a complete bag of s**t. I have seen five axis milling which takes 20 min on the machine record a time of 6 min in SV (and take 4 hours plus to run through in SV) and a gun drill operation that records 1½hrs in SV and takes 2 mins on the machine.

SolidVerify is the weakest thing in SolidCAM and comes from a software house called MachineWorks. MachineWorks supply verification software to lots of CAM companies - have a look at their website.

[mattpatt]

I'm with Bob on this one.

I think that I brought up this subject a while ago. The time shown in any of the verifications should be totally ignored in my opinion. I pay no attention to it.

When I have the code I run it through CIMCO Edit V5, which I've found gives me a much closer time estimate.

Just yesterday I finished a mold I was making. Solidcam reckoned on 23hrs for the finish cut or something silly. CIMCO estimated 3 1/2 hours. Actual time taken was 3 hours 20 mins.

[Brakeman Bob]

CIMCO Edit V5 is a good program. Just wish that their simulation worked in five axis as the simulation in palin 3D is very good. Just wish I could work out how to save a tool library so that I didn't have to redefine the tools all the time.

[pinguS]

Hey all
i just seen times vary between, host cad simulation, solid verify and rapid verify. With rapid verify coming the closest. What an excellent program solidcam is, and then they f&*k you right off with the verification process.
Maybe they should hold back on one years update, and spend time on the verify stuff, and then they will have one of the best cam softwares.

Saying that, i already think its one of the best, so really just venting...

[Brakeman Bob]

With SolidVerify dependent on a third party for development I hold very hope that SolidVerify will be improved as we would like, however, it a lot cheaper than VeriCut and it does a job of sorts as long as you don't want cycle times.

Now what I have been working with lately is getting my post to output a time study when the program is posted (when you generate code choose the "Generate code with time" option). It took a fair bit of work in the GPP file to get things in the format I wanted but I am nearly there now. The times are better than those kicked out bu SolidVerify but still are not accurate enough for quoting purposes so I am embarking on a data collection exercise to get true machine times for each job in the job tree which I will use to compile a table of compensation values for each job type / strategy. This will (hopefully) make the times more reliable.

I found CIMCO Edit was great for processing 3D and 5 Axis code for time but it saw Heidenhain drilling cycles as zero time, so it was useless for my purposes.

[dengo]

I'm guessing that you have already adjusted the time factor in your .mac file ??
I have mine set at 1.2 which is a little closer to my actual cycle times on the machine. But then again I don't recall getting times that were out by 1/2 a day even on all night ops.

[Brakeman Bob]

I have tried the time factor in the MAC file but I want to fine tune my post so that individual operations reflect reality a little better. Oh for a lookup function in GPP!

[mattpatt]

I have very little faith in the SC times, so I tend to just generate the code and look at it with CimcoEdit. Which I have found to give much more realistic figures.

[Brakeman Bob]

Matt I agree with you about CIMCO Edit when you are doing Fanuc (or similar) but for Heidenhain it's well adrift because of the cycles don't register as machining. Pity that.

[MKproto]

Discounting time in simultaneous 4 and 5axis operations, I have tweked mine over the years and get simulation times on 2 aixs part within 5-10 seconds and within a minute or two on 20-30 min 3d cycles I use it all the time. I have a spread sheet of calulated vs actual times for about 120 parts that I have run in the last 3 months and they are pretty much dead nuts as far as a simulation goes.