AT THE HEART OF THE ROAD TRANSPORT INDUSTRY.

Call our Sales Team on 0208 912 2120

Cummins uses a computer as the first assessment of an

13th February 1982
Page 44
Page 45
Page 46
Page 44, 13th February 1982 — Cummins uses a computer as the first assessment of an
Close
Noticed an error?
If you've noticed an error in this article please click here to report it so we can fix it.

Which of the following most accurately describes the problem?

engine. It can simulate a road test in five minutes but is never used as a substitute for good judgment. So CM road tests remain the ultimate before actua operation.

VEHICLE Mission Simulation is not something from American space research, but is the title of a computer program. Known as VMS for short, it is used by Cummins to eliminate a lot of expensive guesswork and testing from specifying a driveline combination.

In a determined effort to make CM's testers Brock and Blakemore redundant, the VMS computer can simulate a Commercial Motor Scottish road test in about five minutes.

But there is obviously more to it than that, as I found when I spoke to Stan Hartshorne, manager, product planning at Cummins' New Malden headquarters.

Stan Hartshorne uses the VMS program quite regularly at about monthly intervals and this is never for just the one run. As the program should only be used for checking one variable at a time, having a minimum number of runs of five is not difficult.

Cummins uses the computer as the first assessment of an engine development. The research data is put into the computer and if the result shows no improvement then the Cummins engineers immediately check the input data to see whether a programming error was made. The experience with VMS is such that the "time wasting" designs and variations are eliminated before any data is fed into the computer.

Only if the computer gives a satisfactory theoretical performance for the vehicle will an actual engine be put into the chassis. VMS is never used on its own. Cummins is adamant that it is not a substitute for good judgment but that it can be very useful indeed in paring down a list of variables before the actual hardware is subjected to a test program.

Cummins has pages and pages of "routes" programmed into the computer and these cover the world, in effect, although the majority are for the USA. Stan Hartshorne has a map of England, Scotland and Wales showing the various roads which have been programmed into the computer for "domestic" use. Thus he can add together various road sections to build up a required route which can in many cases be used to simulate actual routes used by a particular operator.

This does not just cover the UK. It is quite feasible to build up a route from, say, Aberdeen to Athens on VMS.

One feature of the Cummins program isthat all routes are in miles and the fuel consumption results come back in US miles per gallon and so must be converted. A similar conversion is necessary for gross weights or, as is their wont, the

km erica ns work in lb rather than onnes or ton.

The program also includes letails of the type of road urface which varies from good oncrete to a rotten surface. The itter is not for off-road use, /Filch has its own separate rogram. The road surface data ; necessary to define the rolling asistance which can go up to ,000 lb/ton for a really sticky urface.

Apart from details of the route le computer also needs 'formation as to the conditions temperature, wind speed and irection and so on. The :andard temperature is 68°F iegrees Fahrenheit is standard Iroughout) which can be aried, but Stan Hartshorne oes not think it necessary for a K operation.

At one time a standard wind )eed of 10mph SW was used

this has now been dropped. The VMS program manual has 3ge after page of code

imbers related to specific vehicles but these tend to be US oriented, eg code 1328 refers to a Peterbilt. Stan Hartshorne uses code 1219, which covers the "modern forward control" category.

The configuration also needs defining in exact detail by a similar coding system. Code 512, for example, in the Cummins manual relates to 4x2-25 in other wards, a 4 x 2 tractive unit with a tandem-axle semi-trailer.

Obviously, the type of trailer will have a considerable influence on the results and this is taken into account with a variety of variable inputs including flat-bed trailer, boxvan with ribbed sides and rounded corners, boxvan with flat sides and square corners and every other permutation you can think of.

Stan Hartshorne tends to go for a smooth-sided van with square corners as being perhaps the most representative for UK use. As a matter of interest Cummins used to use a flat-bed trailer and switched to a boxvan at about the same time as CM. Practical results tend to agree, too. Typically Cummins has found that a boxvan accounts for about 0.75mpg on the whole route, overall.

The frontal area of the vehicle combination is of tremendous significance especially as the road speed increases, so the next column on the VMS input to be filled in covers frontal area. This is derived by a simple width x height above the ground calculation with a standard gap of nine inches being allowed for the undercab clearance.

The use of a wind deflector can be allowed for by applying a "judgment factor" and using previous experience to reduce the effective frontal area.

The standard drag factor built into the program for a boxvan combination is 0.7. If the vehicle to be tested has a known drag factor of, say, 0.6 then the frontal area data is reduced by a factor of 0.6 divided by 0.7. As Stan Hartshorne put it: "This may seem crude but it gives a good enough result in practice."

As any operator knows, the specification of tyre can have a marked effect on the fuel consumption. This relates not only to the tyre size and its effect on wheel revs/mile but also to the type of construction, ie radial or cross ply. If the factor for a cross ply is taken as unity then the equivalent for a radial design is 0.7. Some of the latest 80 per cent aspect ratio tyres are down to 0.62.

All this may seem very complicated, but it takes longer to say than do. In Stan Hartshorne's experience, it takes him five minutes to fill in the relevant details on the input form but then "I've done my homework first."

The details for the vehicle specification are relatively easy to come by, however. It is the profiling of the actual route which takes the time, with the one advantage being that once it has been done, the information is in there for life. The CM route (the old one, starting at Hemel Hempstead — Cummins has not yet got around to surveying the latest one) which covers every change of gradient takes up 36 pages with 24 lines of data on each page. The CM route as visualised by the VMS computer is shown in the accompanying map.

A very important factor in determining the fuel consumption for a new driveline combination is the cruising speed. Cummins is now moving gradually towards a higher maximum geared vehicle speed to keep the engine speed in the maximum efficiency sector when cruising at the 60mph limit. Currently the VMS system uses 55mph as a European mainland limit and 60mph for the UK.

This subject of speed limits is an interesting one as it does presuppose that drivers always keep to the legal limits. To allow for this, Stan Hartshorne has two code numbers for the CM route. One covers a strictly legal driver while the other gives more leeway on road speeds. Thus the motorway speed limit is taken off and the Al is programmed for 45/50mph instead of the legal 40.

This policy has proved extremely useful when investigating the effect of roadspeed governors.

Traffic has always been a problem with road tests in that vehicle A might have a clear run whereas vehicle B in the same weight category might run into very heavy traffic. Over a 700plus-mile route such delays tend to balance out, but this is difficult to achieve in a computer program. Cummins puts in an arbitrary speed limit to simulate traffic conditions in that if the legal maximum speed through a town centre is 30mph then VMS estimates an average speed of 18mph.

The conventional highway VMS program has stored input data for most known mechanical transmissions. Automatic transmissions are not included as these are mainly confined to off-road dump trucks for which Cummins has a separate program.

When the button is pressed, the computer goes to work with the whole "road test" being accomplished in some five minutes, for which Stan Hartshorne is charged at a flat rate plus X cents per mile.

The first results printed are a repeat of the input data ie what the operator actually fed into the computer as opposed to what he thought he put in. Thus if any odd results are thrown up, it can be checked that this was or was not due to an incorrect input. The computer then prints out the theoretical performance eg geared maximum speed and gradeability as though the calculations had been made, with a slide rule (but a lot quicker). However, that is only part of the story.

What the computer can also do — and again in a matter of minutes — for the specified combination of vehicle and route, is to calculate how much time was spent in each gear, how many upshifts and downshifts were made, the average revs/mile, the total time at full throttle and so on.

All this is in addition to the fuel consumption and the average speed for the chosen route.

VMS will never be used on its own. There is no substitute for testing the hardware under actual operating conditions. What it does do is provide a quick and accurate guideline as to the results likely to be achieved by a vehicle in practic Where it saves a lot of time anc money is on transmission ratio optimisation whereto test evei possible combination of axle and gearbox ratio would be a long and tedious job.

Current performance shows the VMS program to be slight!) pessimistic. As Stan put it: "It': time we had a better driver in t computer." The program agreo in relative terms, ie in the ordel of percentage improvement ar is only slightly out in absolute terms.

Apart from being used for Cummin's own use, Leyland at Ford have used it in the past as have some of the larger fleets. Although Cummins will use VMS to determine a spec for at operator it is really only for the larger operator for obvious reasons. So don't immediately get on the phone to Cummins!

The VMS computer is based the Cummins research and development establishment at Columbus, Indiana, which is n, as inconvenient as it sounds because the said establishmer is linked to the Cummins facto at Darlington. What Stan Hartshorne does is to phone h input data through to Darlingt( where the operator will punch out to Columbus by direct link has happened that the print-o of the results has come back L the following morning, but thi is rare as Columbus itself use the computer frequently. For Cummins UK, three days is th norm.

Thus it is possible to carry c a full 700-odd-mile road test ir about five minutes. Eat your hearts out Brock and Blakemc

Tags

Organisations: US Federal Reserve
People: Stan Hartshorne
Locations: Aberdeen, Athens, Columbus

comments powered by Disqus