Will This Machine Work?

Find Best Test Methods for Proving Your Machine Operation

1 of 4 < 1 | 2 | 3 | 4 View on one page
Virtual design and prototyping are a controls engineer's best friend. Find out how the Control Intelligence Agency is leveraging that expertise. Watch the "CIA Virtual Brigade Briefing Trailer".

By Dan Hebert, PE, Senior Technical Editor

There are many different ways to test and verify that your machine and its automation system will perform as intended prior to final fabrication, installation and commissioning. Each test method has advantages and disadvantages, and different machine builders employ disparate testing methodologies depending on particular circumstances.

The main testing methods are software emulation, hardware simulation, prototyping and beta testing at a customer site. Many machine builders mix and match these test methods and also employ different testing protocols for different machines.

Software emulation just tests out the controller and HMI software programs. This entails writing a program to test the basic on/off and dynamic operation of the machine. Dynamic operations such as control of movements either are not simulated or are simulated in a simple form without detailed analysis.

Hardware simulation starts with a detailed and accurate software-based model of a machine’s hardware including dynamic operations, clearances and interactions among moving parts. A good hardware simulation takes lots of time and effort to develop but in the best case can simulate virtually the entire operation of the machine and even parts of the customer’s factory.

Table 1:

Test Methods from
Simplest to Most Complex

1. Software emulation
2. Hardware simulation
3. Prototyping
4. Customer site beta test

Prototyping involves building all or part of a machine and its automation system. Prototype machines are tested in the shop to ensure correct operation. This is a step up in time and expense from hardware simulation. But many feel prototyping achieves better results than simulation.

A customer-site beta test requires installing a completed machine at a customer site and then testing the machine in actual production with all or part of a customer’s manufacturing line. This is the most expensive type of testing and requires a compliant customer, but there is no better way to verify machine operation.
We’ll examine some of the pros and cons of each test method, starting with the simplest and progressing to the most complex.

Software Emulation

Software emulation is the most basic form of testing and is often sufficient for standard machines that have been in production for some length of time. “We use a combination of prototype parts and software emulation for machine testing,” says Paul Brancaleone, principal controls engineer at plastics processing equipment builder Gloucester Engineering in Gloucester, Mass. “Before software actually makes it to a machine, we simulate and test the response and sequences. We build debugging code into our programs using our source assembly language,” adds Brancaleone.

Software emulation is most useful for standard machines, but many machine builders find that these machines work well enough to skip the step and go straight to shop test. “We carry out a factory acceptance test on each machine with the customer before it leaves our manufacturing facility,” explains Ken Carter, general manager and president of EDL Packaging Engineers in Green Bay, Wis. EDL makes end-of-line packaging machinery.

“Emulation software for the PLC program might be useful, but I am not sure how much time it would save. It is fairly easy to modify the PLC program during the shop test if there are errors and if modifications are required,” adds Carter.

Another machine builder seconds Carter’s point. “We only spend 80 hours or less to program a machine prior to shop setup, so software emulation isn't necessary for us,” claims Wayne Kornelsen, supervisor of controls engineering at Priority One Packing in Waterloo, Ontario, Canada, a maker of palletizing equipment.

“If we needed to develop a new machine from scratch, we would need something that could simulate cycle times and both pneumatic- and drive-driven motion. One thing our company could use is a low cost simulation package for cycle time and throughput estimating. We have looked at Arena from Rockwell Automation and Simul8 from Simulat8 for this, but they are fairly expensive.”An example of this was an expander turbine start-up Maverick did a few years back. “When we got to the site the PLC’s ControlLogix program was incomplete and obviously untested,” says Gellner. “ We were not responsible for the PLC programming, but because of the propensity of the expander to upset the process, which could result in explosions, we set up a SoftLogix controller, loaded the PLC program, added some very extensive loop-back code, connected it to the HMI and shook out the wrinkles before we put it online.” Explosion avoided.

1 of 4 < 1 | 2 | 3 | 4 View on one page
Show Comments
Hide Comments

Join the discussion

We welcome your thoughtful comments.
All comments will display your user name.

Want to participate in the discussion?

Register for free

Log in for complete access.

Comments

No one has commented on this page yet.

RSS feed for comments on this page | RSS feed for all comments