The engineers at Ohio-based INVOTEC have another term for mechatronics, engineering’s latest buzzword.
“We just call it work,” says engineer John Hanna, president and co-founder of the Dayton area design firm, which employs some 60 engineers, technicians and machinists. Ever since Hanna and engineering colleague Daryl Greywitt launched INVOTEC in 1993, the company has made it a practice to blend mechanical, electronic, controls and software programming skills in its design approach.
Because of that integrated skills set, INVOTEC has garnered a steady stream of turnkey automation projects for customers in industries ranging from automotive to medical to materials handling. The technologies that the firm’s automation cells incorporate include: ultrasonic welders, adhesive dispensers, machine vision, force and torque measurements, mechanical assembly and much more. Virtually every project must provide a motion control solution and requires a thorough understanding of the customer’s overall manufacturing operation.
Recently, Design News asked one of INVOTEC’s teams to share their insights on a recent mechatronics project, a laser-based material removal cell for a medical component company. Their observations show that the constant give and take among team members representing different engineering disciplines is every bit as important as the components chosen for the automation solution.
Design News: Before we get into the laser material removal application, tell us what makes INVOTEC different from other design firms or system integrators?
John Hanna (INVOTEC president): Our niche centers on accomplishing assembly and test tasks that require a high level of controls engineering sophistication, integrated with a well-designed mechanical approach. Most of our systems require integration of several different technologies from a variety of vendors. We also tend to be successful with smaller parts that are typically more difficult to handle. So the high-precision automation and test cells that we design typically measure about 4 ft by 4 ft, although one assembly system project we designed and built joined several stations together into a system that measured 40 ft long by 20 ft wide. Due to the niche we have established, we employ as many electrical engineering experts as we do mechanical engineering experts. The nature of our work requires our engineers to maintain a high level of sophistication and competence with a variety of systems and vendors.
We also don’t shy away from or subcontract work that requires integration of multiple platforms and technologies. And we manufacture our designs, providing them to the customer as complete systems. By depending largely on our own internal resources, we are better able to control the process and provide a system that functions with a high level of reliability.
Design News: What were the major design challenges in this laser-based materials removal application – and why did the customer come to INVOTEC?
Daryl Greywitt (company VP and project leader): We were selected for this project because of our willingness and demonstrated ability to think creatively about a problem and arrive at a solution based on the constraints that the customer outlined: cost, timing, reliability, serviceability. Laser processing was not the first and only solution that we evaluated. Nor did the customer come to us and ask for a laser material removal system meeting a given specification. They came requesting the best solution for de-gating and removing flash from an injection-molded, synthetic-rubber part. The part was extremely elastic in nature, and the required final shape was irregular and held to very tight tolerances. In addition, the customer wanted a total cycle time of about eight seconds for the entire operation, which includes placing a nest of parts into the station, laser cutting the parts, and removing the nest from the station. Before embarking on the laser solution, we evaluated and even prototyped several technologies, including mechanical removal with a die or cutter, water jet cutting, and ultrasonic cutting.
Design News: What were the most important mechanical components that you chose for this design?
Todd Koewler (lead design engineer): The key to getting the throughput and accuracy we needed was XY positioning slides for the cutting system. We used Parker 404T series linear motor slides. When we were deciding what we should use we took several factors into consideration. The movements we needed to make were high speed and high acceleration, which led us toward the linear motors. The accuracy on the slides we used were 5 micron to accommodate the tight tolerances for the parts. An added benefit of using these slides was low maintenance and low friction.
Another big mechanical challenge was the support structure for the overall system. The high speed of the linear motors required that the base of the system have enough mass to withstand the accelerations. We also had to support a rack-mounted cooling unit for the Photonics UV laser, which was attached to the side of the machine. Surrounding the entire work area was an extruded aluminum shroud to contain the UV laser. Access to the working area was needed for a variety of cords, cables, and hoses, as well as a sliding door in the front to load the parts onto the nest for cutting. We also had to devise a hose and ductwork system to remove the fumes from the laser-cutting process.
The bottom line was that this system needed to cut good parts without damaging them. There was nothing our equipment could do to make a scrap part good, but we had to ensure that the system did not malfunction and ruin otherwise good parts.
Design News: How about the electronics and controls challenges on this project?
Michael Updike (controls engineer): Key components from a controls standpoint were the Parker 6K2 motion controller and a Cognex InSight 5403 high-resolution camera, which was mounted above the slides. The camera’s job was to inspect the part and send out corrections to the laser-cutting path, stored in the motion control system. I had to develop the software program that took the coordinates acquired from the camera and communicate them to the linear slides. The camera also identified defective parts that would not be cut and therefore scrapped at this station. An Allen Bradley MicroLogix 1500 PLC acted as the overall sequencer, triggering the camera to send the commands to the motion controller. The PLC triggered the camera to set up the calibration target, move the power meter, as well as inspect the part. The operator interface for all of this was an Allen Bradley Panelmate touchscreen.
Perhaps the biggest challenge in this operation was gaining speed. The original interface between our machine vision system and our motion control system was serial. It took about 10 seconds to pass all 125 variables between the two systems. The solution was to establish an Ethernet connection between the camera and the drive. I didn’t have instructions on how to do this, but I did have some instructions on how to make the link to a DVT camera. I had to format the messages like a DVT, and trick the Parker into opening the connection. All this cut the communications time for the data to about a second.
Design News: What were the chief design tools that you used on the project?
Todd Koewler: We used the Autodesk Inventor 3D modeling package for the overall design of the station. Solids modeling is a tremendous help in putting the overall machine through its motions and evaluating the performance of the slides. This software package also has built-in finite element analysis capability, which was useful as we sought to minimize vibration in the system. We also did 2D AutoCAD work to help create the ideal cutting path on an ideal part. We had to measure all the angles for the path and get the right radii – and we had to get this information from the right reference points so that Mike could use them in the Parker controller. There were many iterations of the ideal path, different radii and different lead-ins for the curves to minimize the accelerations and decrease cycle time. We had to spend a lot of time optimizing the cutting path, given the constraints of this elastomeric material. In general, our design projects feature systems that are increasingly controlled with personal computers and that run applications such as LabVIEW for system control, measurement, data collection and storage.
Design News: How about tests to prove out the design?
Michael Updike: Much of the time we used the system itself to test out various aspects of the system. I would have the laser trace a pattern on the calibration target and then use a secondary camera looking down the laser optics to see if I could reliably trace the same pattern from the same starting position. We used many different program iterations in both the camera and the motion controller to test out different factors, such as how soon to start slowing down the slides after a long transition run before we start the rapid intricate turns that we still need to do fast. If we slow down too soon, we add time to the operation. If we slow down too late, we risk position error.
Design News: How important was it for the various engineering disciplines to share their ideas and design plans?
Michael Updike: As the controls engineer on this system, I was doing most of the testing for vision, motion, and laser operation, and I depended on the designers to make new tools for me. For example, we had to devise custom lights to backlight the parts through the removable nest. We had a couple iterations to get the red LEDs positioned correctly inside four white Delrin blocks so that the cutting area would be lit evenly. That was a great example of mechanical and controls functions working together.
Another challenge was integrating our physical laser shutter. We had to have a physical barrier in place for safety reasons with the laser. The sliding access door in the front has a locking safety switch linked to a physical beam-dump shutter on the laser. This prevents the laser from operating when the access door is open.
Design News: What was the overall result of the project – and the customer’s role?
Daryl Greywitt: Engineers from the customer company were involved every step of the way, since we were in effect developing the specs on the fly. The customer pushed the technology envelope to get the system to perform faster and to tighter specs. In all, it took us about 18 weeks to complete the project, including some of the preliminary evaluations of different cutting systems. In the end, this design allowed the customer to get their product on the market faster, as well as meet the cycle time goals they wanted for volume manufacturing.