Technology news
Multipoint PLC-Based Control Optimizes Bump-Tubing Extrusion
To maintain consistent dimensions in bump-tubing diameters during extrusion, producers of medical tubing vary the internal air pressure, process speed, and other extrusion parameters. This has traditionally been the function of 
timers and switches, an alchemy that Charles Sparacino, sales engineer at Davis-Standard Corp. (Pawcatuck, CT, USA), dubs a "black-box art." Davis-Standard's incorporation of a 25-point control system that communicates with a host computer brings the process within the realm of engineering science, according to Sparacino. Specifically, it enables tighter tolerances, decreased production costs, and process validation.
The production of bump tubing brings to the extrusion process the complexity of forming and sizing more than one diameter as the part is extruded, says Sparacino. To accomplish this, "the puller speed is lowered while the internal air pressure is increased to create the larger diameter. Then the process is reversed to bring back the tube to the smaller diameter," he explains. "In order to maintain a consistent size of either or both diameters, the rate at which this is done must be accurately controlled." In addition, the tubing cutter must obey precise parameters and the puller must change speeds at a constant rate. Cumbersome timer and switch control systems, according to Sparacino, limit the range of products that can be manufactured, as well as production throughput.
"The microprocessor has changed this method by simplifying the user interface and enabling the puller and cutter control to communicate directly with the puller drive, air control servo, and cutter," says Sparacino. The company's controller, he adds, offers 25 adjustment points per axis to ensure smooth transitions. "We could easily increase the number, but we have found that for user applications 25-point control is sufficient."
Once the process parameters have been set, they can be downloaded with one recipe command. All of the component parts of the system can be monitored and controlled from a single operator control station.
Moulding Process Slashes Production Cost of Teflon Flowmeter
A proprietary disposable flowmeter made of PFA, or Teflon, was honored at the recent Salon International des Inventions in Geneva. Equflow Sensoren (Wychen, Netherlands), which developed the Teflon Click Sensor, received a silver medal at the event for its moulding technology that enables the cost-effective production of small PFA parts such as the sensor.
All of the sensor's wetted parts, with the exception of the ruby bearings, are made of PFA. The component can withstand sterilization temperatures up to 160°C. "Sterilization is simply a matter of removing the flow pipe with the metering device from the [clip]," says Equflow's technical manager Fred Wooninck. The sensor, which can measure all coloured liquids, achieves dosing levels starting at 0.1 ml with a repeatability of ±0.15%.
"When people hear 'Teflon sensor' they immediately think high cost," says Wooninck. The production of a small PFA part traditionally requires machining, which significantly increases production costs. "Certainly, larger parts can be moulded in PFA, but until now it could not be done with items on the order of this 4.5-mm sensor that weighs 0.015 g. We are the only company currently able to mould this part in PFA," he says. Because no machining is involved, the company can offer sensors such as this one at substantial savings, he adds.
Flow sensors are available with threaded, hose-barb, and custom connectors; a variety of electronic controls can be supplied for easy integration of the sensors into flow-monitoring and dosing systems. The company also offers these sensors in fibre-optic versions, and welcomes OEM requests for custom components.
Classifier Adds Fuzzy Dimension to Quality Control
An object-oriented system speeds the design and development of turnkey and embedded visual object-recognition solutions. Created by Braintech Inc. (North Vancouver, BC, Canada), the Windows NTbased Odysee development system features the Braintron fuzzy classifier—a sensor-independent module that allows the system developer to combine or fuse information from different data streams.
The Braintron fuzzy classifier reportedly provides all the advantages of neural-net learning, as well as an additive function that enables training, retraining, and editing on the fly. Well suited for the rapid development and testing of automated medical imaging systems and other custom machine-vision applications, the classifier implements the company's patented technology based on a set of highly adaptive pattern-matching algorithms.
According to research and development manager Charles Hooge, the classifier is a robust image classification tool that enables system developers to focus on data acquisition and preprocessing rather than image recognition. "The advantage," says Hooge, "is that each new class is its own fuzzy signature and [the classifier] spawns fuzzy signatures as you work." For example, says Hooge, the system developer shows the classifier rivet A and, from that point, it recognizes rivet A. Then the developer shows the classifier rivet B and, from that point, it knows rivets A and B. A set is formulated in such a manner.
This building of the database over time, explains Hooge, makes the development system an extremely stable one. "A medical device manufacturer who is looking for defects," says Hooge, "can build continuously a data set that will [help to pinpoint] them. Then, the manufacturer can see the deterioration of the quality of the instrument before anything actually fails, and go back into the quality control process to make any necessary changes. It's preventive maintenance of your production process."
Other key features of the Odysee include a drag-and-drop tool to create vision solutions, remote collection of data, and the ability to change hardware components such as frame grabbers or cameras without reengineering. Each subtask of the system&emdash;—acquisition, digitization, preprocessing, classification, and database maintenance—is performed within a self-contained component. Benefits of the system include expandability, interchangeability, and the ability to reuse modules in new applications.
The firm has recently completed a feasibility study for an automated stent inspection system.
