Windmill Software Ltd
Windows Engineering Software

Monitor Newsletter Archive

Home
Newsletters
Shop

Issue 12: Product Reliability Testing; Engineering Newsgroups

CONTENTS

Windmill News | Product Testing | Internal Combustion Engines | Switch Testing | General Principles | Advantages of a Computerised System | Engineering and Scientific Newsgroups

WINDMILL NEWS

In August we're implementing a library of instrument setup files. Many thanks to those of you who have already sent in your files.

With our free RS232 instrumentation software you can connect almost any ASCII RS232 instrument to a Windows data acquisition system. However, you have to tell the software the configuration details of your instrument, and each type of instrument needs configuring differently according to the manufacturer's instructions. Our library of files means that if someone else has already used Windmill with your instrument, you can download the file and get started immediately.

If you would like to contribute please e-mail your *.aid file, *.ims file and details of your instrument to techsupport@windmill.co.uk. (The files are text files and only contain information about your instrument. To check the file contents just open them in Notepad.) If you wish we'll place a link back to your web site in the library, as an acknowledgement of your help.

PRODUCT RELIABILITY TESTING - PRINCIPLES AND PRACTICAL EXAMPLES

Continuous, long-term, testing ensures the safety and suitability of electrical and mechanical products. Round-the-clock monitoring, with automatic shut down should a danger level be crossed, is a common task for an unattended data acquisition system. For the majority of products the same general principles apply. Here we illustrate those principles with 2 specific examples. Let’s look first at internal combustion engines and then at electric light switches.

*

Internal Combustion Engine Testing

Durability and Performance Testing
The test procedure runs the engine slowly for part of every hour, and at full speed for the rest of the hour. The timings simulate the typical use of the engine: a tractor engine might be idly ticking over for 55 minutes and at full speed for 5 minutes for example. This is called the duty cycle and measures durability (tickover) and performance (full speed).

Making the Measurements
During the length of the test, which is often several months, a range of measurements are taken. These include fuel consumption, fuel temperature, top hose temperature, oil pressure, torque, exhaust back pressure, engine speed and smoke test.

Transducers convert signals from the engine into a form the data acquisition hardware can understand - normally voltage but sometimes current. For example, to measure pressure, an air or oil hose runs 3 or 4 metres from the engine to a pressure transducer. The transducer has a diaphragm that converts the pressure into an electrical signal we can measure.

Temperatures are often measured with K-type thermocouples. These are robust and cheap, and sufficiently accurate for the task. For a discussion of thermocouples see Issue 5 of Monitor.

Speed is measured in revolutions per minute by a tachometer. This produces either a voltage output, or a pulse output. You need a counter interface module to handle the pulse output.

You wire the transducers to the data acquisition interface. This is often remote from the computer and connected to it over an RS485 cable. RS485 allows long distances between the test site and the PC (around 1000 metres) - protecting the PC from unsuitable environments. It also lets you distribute several data acquisition devices around the site, connecting them all to one cable and computer. It is quite a slow communication link (handling around 35 samples per second), but for product testing such as this high speeds are not needed.

Software Control
How often measurements are taken from the engine, and what happens if a measurement exceeds its danger level, is controlled by the software. If, say, 6 engines are being tested by one PC, the software typically logs around 200 measurements every minute. Once an hour the software tells the engines to increase speed to maximum revolutions. It does this by sending control signals down analogue output channels. After a specified time it tells the engine to return to tickover speed. The software provides permanent records of how key parameters have changed, highlighting any long term drifts. Importantly, the software shuts the engine down should, say, oil pressure be lost - saving the engine from irrevocable damage.

*

Switch Testing

Our second example is testing electric light switches, like those on the walls of your house or office. We measure the temperature of a switch, making sure it doesn't get too hot, and the voltage drop across the switch. Ideally the temperature would remain constant and the voltage at both ends of the switch would be identical. In practice, because contact wear and switch oxidation increase resistance, the temperature will increase and the voltage drop.

Durability and Performance Testing
A pneumatic finger, under software control, mechanically presses the switch on and off. The duty cycle of switch testing is typically: switch on for 4 seconds and switch off for 6 seconds. Whereas in our engine testing example the duty cycle lasted an hour, in switch testing it lasts for 10 seconds. This cycle is repeated 30 000 times.

Periodically, at 25%, 50%, 75% and 100% through the test, the software leaves the switch on for an hour and records the rise in temperature and voltage drop across the contacts.

Like in the engine testing example, the temperature is measured with a thermocouple. In this case the thermocouples are mounted on the contacts at mains voltage, and so must be isolated.

Automatic Shut Down
As well as counting the number of pneumatic finger operations, the software counts the number of voltage state changes (on-off). These two counts should be identical, ensuring that the arm is not missing the switch and that the switch is still functioning. If the counts differ, indicating that the switch has not been switched on or off, the software automatically stops the test.

It will also stop the test for safety reasons - should the temperature become too hot for example.

Permanent Records
The software records all the measurements, counts and alarms. These can be automatically passed, to, say, a spreadsheet for analysis and a copy printed suitable for the standards inspectorate.

*

General Principles of Product Reliability Testing

Although we have discussed two very different products, the method of testing is very similar in each.

1. The tests are controlled by software, with automatic shut-down should alarm levels be crossed.
2. The software saves permanent records, suitable for proving the reliability of the product to government standards organisations.
3. The tests are 24 hours a day, long term (around 6 months for engines) and unattended.
4. The tests each have a "duty cycle": operating at different states over a period of time. One hour for engines and ten seconds for switches.
5. Durability and performance are both tested, either as part as the duty cycle (engines) or by periodically interrupting the duty cycles (switches).

*

Advantages of a Computerised System

A computerised data acquisition system brings considerable advantages to product testing. As the tests run continually it saves a great deal of time. It also relieves people from monitoring the test, saving many months of labour. It is consistently accurate - all readings taken at the same time with no transcription errors. It can automatically generate comprehensive reports, on any computer on the network that you choose.

For more details of product reliability testing, or to discuss your application, please contact Windmill Software at sales@windmill.co.uk.

ENGINEERING AND SCIENTIFIC NEWSGROUPS

The internet has brought us an unlimited supply of free advice - what with world wide web pages, bulletin boards, chat rooms and the like. An invaluable source is the Usenet collection of newsgroups.

Usenet stands for Users' Network. It has been around some time - having been started in 1979 at Duke University and the University of North Carolina. Usenet comprises thousands of newsgroups, each covering a specific topic. People around the world can post comments and questions to a newsgroup by e-mail. Chances are that whatever your problem, someone, somewhere, will already have addressed it and can offer either a solution, or at least a pointer of where else to look.

You first need to find out though, which newsgroups cover your industry or interest. The first clue is in the title. All newsgroups are named according to a hierarchical classification system, hierarchies being separated by dots. For example, sci.engr.control. The first syllable, sci, indicates that this is a scientific newsgroup. Other examples of top level hierarchies are comp (computer), bionet (biology network), alt (alternative), eunet (European network) and misc (miscellaneous). For a full list see http://www.magma.ca/~leisen/mlnh

The second syllable, engr in our example, indicates that the group pertains to engineering, and the third to control. The first part of the name is the most general (sci), the second more specific, and so on.

Your internet service provider generally carries only the most popular newsgroups. There are some places on the web, though, that give more comprehensive lists.
http://tile.net/news/
http://www.metalab.unc.edu/usenet-i/hier-s/top.html
ftp://ftp.lsc.org/pub/usenet/CONFIG/newsgroups
(The last resource is very large and therefore slow to load.)

Another way to find the newsgroup best fitting your interest is to search the Usenet messages that have been posted. The most popular newsgroup search engine is at http://www.deja.com/. Enter your search terms and choose to search "discussions".

A third option is to ask in the newsgroup news:news.groups.questions Someone will then tell you which newsgroup is the most appropriate.

Once you've identified possible candidates it's a good idea to read their FAQs. These serve two purposes. They tell you what is and isn't appropriate for posting to the group. They also list those "frequently asked questions": saving you time and making sure you don't irritate group regulars. Collections of newsgroup FAQs are at
http://www.cs.ruu.nl/cgi-bin/faqwais
http://www.faqs.org/
http://www.metalab.unc.edu/usenet-i/hier-s/top.html

Finally - a few examples of the scientific and engineering newsgroups available.

news:alt.engineering
news:alt.engineering.electrical
news:bionet.software
news:comp.home.automation
news:comp.robotics.misc
news:misc.books.technical
news:rec.autos.sport.tech
news:sci.electronics.basics
news:sci.electronics.equipment
news:sci.engr
news:sci.engr.chem
news:sci.engr.control
news:sci.engr.electricalcompliance
news:sci.engr.hydrodynamics
news:sci.engr.manufacturing
news:sci.engr.marine
news:sci.engr.mech
news:sci.engr.safety
news:sci.environment
news:sci.materials
news:sci.misc
news:sci.techniques.spectroscopy

A much longer list of engineering groups is provided by the Department of Engineering Science at the University of Oxford .

Copyright 1999 Windmill Software Ltd. All rights reserved. This newsletter may be distributed in its entirety. However, individual sections may not be reproduced without the prior written agreement of Windmill Software Ltd.

Subscribing

To subscribe to Monitor please fill in your email address below. We will not pass your address to any third parties.