Board logo

Front upright kit
v8kid - 17/4/10 at 05:04 PM

Some time ago there were extensive discussions on a custom front upright kit mainly lead by Rorty if I remember right.

Is there still an interest in lightweight kits of parts for members to assemble their own uprights?

I have been working on Rorty's design for myself and have modified it quite extensively stiffening it in buckling, bending and torsion whilst not adding weight.

The advantage of a weld it yourself kit of parts are that the location of the lower and upper pick up points can be determined to suit yourself - handy if you are running extra high or extra low ride heights. Another advantage is that the kpi and ackerman angle are easily adjustable at the welding stage.

What range of adjustments would members like to see? I am very content with cortina geometry - around 5 degree kpi and 80mm scrub but aware that we all have our own directions to follow.

Cheers

David


russbost - 17/4/10 at 05:33 PM

Any links to previous discussions?


cymtriks - 18/4/10 at 12:20 PM

The Elise has:
KPI 12.0 degrees
Castor 4.25 degrees
Trail 4mm
Scrub radius 10.5mm

The MX5 (Miata) has zero scrub

Mass makes have been using high KPI and low scrub for a long time. Why on Earth would you want to copy anything from a Cortina? I've heard that parts are getting harder to source to.

Why did you feel the need to change Rorty's design?


iank - 18/4/10 at 12:41 PM

quote:
Originally posted by russbost
Any links to previous discussions?


http://www.locostbuilders.co.uk/viewthread.php?tid=33544

http://www.locostbuilders.co.uk/viewthread.php?tid=34565


v8kid - 18/4/10 at 03:59 PM

Thanks for posting the links Ian. It all started with this article which addresses the subject of ideal spindle geometry. The questioner postulated that zero scrub and zero kpi seemed a good idea and Mark Ortiz made the case against it.

www.auto-ware.com/ortiz/ChassisNewsletter--January-February2005.

To paraphrase it and add my own interpretations:-

Steering with little or no trail will go light before the point of tyre breakaway – excellent for a commuter vehicle not so hot for a sports car. Trail also improves turn in. So we need some trail.

A car with little or no scrub radius tends to be numb and uncommunicative so we need some scrub please. Commuter cars have very little because it increases vibration and braking feedback through the steering wheel – I think car engineers call it NVH and it is paramount to the sales chappies - even in the Elise. Scrub also helps to unload the rear wheel in tight corners – handy if you have a LSD

The camber change associated with caster is good and caster combined with scrub helps turn in for slow corners particularly in cars with LSD so a dollop of that as well please

The kpi’s used in modern cars are more due to packaging considerations necessitated by MacPherson struts which in turn are required to get widely spaced mounting points on a pressed steel frame. They also make up for inattentive drivers and none of the above applies to us does it! So not so much kpi for us.

All in all a persuasive case for not adopting suspension uprights from heavy saloon cars designed for commuting rather than spirited driving. It is hardly surprising however that the humble Cortina has some of the attributes we are looking for – it did not have strut suspension, power steering or any compromises forced on it by NVH requirements. Just because it is old does not justify decrying a design based on sound engineering principles and constraints.

The question is exactly what values to use and this is where an upright kit that could easily be configured to individuals requirements comes it.

What was wrong with Roarties design? Nothing it was perfect for an off road buggy with lots of kpi and zero scrub to reduce steering feedback on rough surfaces. We want an upright for a road sports car.

Also the design was a little weak in torsion and it was difficult to get the bottom outer balljoint far enough outward – ideally we want it just clearing the disc. Also the balljoints were vertical and its easier to package them close to the rim (bottom outer balljoint again!) if they are horizontal.

So that’s where I am coming from and am interested if this is a common problem and if so what range of values people would be looking for. On the other hand it may be just a bee in my bonnet.

Cheers

David


Ahdammit - 19/4/10 at 04:23 PM

Didnt the original 7 use the Triumph upright?
I dont know the geometry or dimensions,but if it saved re-inventing the wheel,what about using the same scrub/kpi but redesigned for modern fittings?


iank - 19/4/10 at 05:11 PM

quote:
Originally posted by Ahdammit
Didnt the original 7 use the Triumph upright?
I dont know the geometry or dimensions,but if it saved re-inventing the wheel,what about using the same scrub/kpi but redesigned for modern fittings?


Caterham still do iirc.


ettore bugatti - 19/4/10 at 09:04 PM

Caterham use only the geometry of the Triumph upright. They changed the bottom trunnion for a balljoint (if Im correct).

A nice alternative could be the Mustang II upright. The hotrodders in the USA use them, widely available and the KPI is 11-12 degrees.


Neville Jones - 20/4/10 at 08:58 AM

The Wilwood uprights are good value, and geometry is useable.

The axles need reshaping for RD hubs, or custom hubs for UK bolt patterns.

Cheers,
Nev.


cymtriks - 20/4/10 at 09:19 PM

quote:
Originally posted by v8kid
A car with little or no scrub radius tends to be numb and uncommunicative so we need some scrub please. Commuter cars have very little because it increases vibration and braking feedback through the steering wheel – I think car engineers call it NVH and it is paramount to the sales chappies - even in the Elise.

The kpi’s used in modern cars are more due to packaging considerations necessitated by MacPherson struts which in turn are required to get widely spaced mounting points on a pressed steel frame. They also make up for inattentive drivers and none of the above applies to us does it! So not so much kpi for us.

All in all a persuasive case for not adopting suspension uprights from heavy saloon cars designed for commuting rather than spirited driving. It is hardly surprising however that the humble Cortina has some of the attributes we are looking for – it did not have strut suspension, power steering or any compromises forced on it by NVH requirements. Just because it is old does not justify decrying a design based on sound engineering principles and constraints.


The Elise :
KPI 12.0 degrees
Castor 4.25 degrees
Trail 4mm
Scrub radius 10.5mm

The MX5 (Miata) has zero scrub

The Honda S2000 has minimal scrub and high KPI but I don't have the numbers.

Are you seriously suggesting that These cars, and many other very fast and capable machines, have got it all wrong?

None of the cars I've mentioned have high KPI to accomodate strut suspension as they use double wishbones.

On the other hand your claim that the Cortina upright is based on sound principles and modern stuff is compromised by catering for struts is just not right. Cortinas were not noted for handling and were designed to cater for the fleet and family workhorse market from the outset. There is no evidence at all that the geometry was chosen to give some double wisbone handling perfection, or even anything other than good use of available tooling and parts.


Livio - 21/4/10 at 09:24 AM

I think one should take in consideration which parameter influences which properties for designing an upright.

-The bigger your scrub the heavier to steer the wheel in a corner. – this must be an important issue in a car without ps. At the same time with increasing scrub the load of the steering mechanism will increase too.
-The more kpi you have the more +camber will be gained by steering the wheel. –does not sound very good
-Trail helps self aligning of course.
-With caster you can gain –camber by steering the wheel.

Usually the limit of reducing scrub and kpi is defined by the brake disc and other packaging constraints. The ET of the wheel has heavy effect on these parameters too.

Of course every parameter is in connection with many other properties and parameters and I just wanted to mention a few from the most important relations.

Correct me if I am not right.

I am curious about the new design.


cheapracer - 21/4/10 at 04:58 PM

Too much about a lot of things that make very little difference except to your Mum's shopping car.

For those who think modern cars are the ducks nuts got to a Historic sedans race meeting and vice versa.

For those who think 5 degrees KPI is great I'll show you winners with 15 and vice versa.

Same for caster.

Trail is the most underated yet very important part of the whole equation for feel and more.

Often when people add caster to their cars they think it handles better for the wrong reasons, it's usually the trail gain thats giving them better feedback and earlier slip angle saturation.

The fastest current LMP2 car in the world uses 0 scrub, 0 caster, 0 KPI, static camber to suit the day but runs a whopping 65mm of trail.

For a road car that doesn't run on silky smooth race tracks and needs suspension travel I would suggest a 2 to 4 degrees KPI , double that in caster, negative scrub of around 20mm +- 10mm and about 30 - 40mm of trail.

neg scrub is pretty easy to get these days with most wheels being for front wheel drive cars with plenty of offset that way.

By the way, Cortina uprights are available in aluminium.

Theres one way a guy did his here...I think the measurements are in the thread somewhere....
http://www.locostusa.com/forums/viewtopic.php?f=36&t=5850&start=15


cheapracer - 21/4/10 at 05:18 PM

quote:
Originally posted by Livio
Correct me if I'm wrong...

-The bigger your scrub the heavier to steer the wheel in a corner. – this must be an important issue in a car without ps. At the same time with increasing scrub the load of the steering mechanism will increase too.
-The more kpi you have the more +camber will be gained by steering the wheel. –does not sound very good
-Trail helps self aligning of course.
-With caster you can gain –camber by steering the wheel.





More scub gives lighter steering, scrub allows the tyre to follow a radius rather than screwing rubber on the spot.

KPI gives you positive camber gain but KPI is a shimmy damper, stability and self aligning giver as well.

Trail does but it does a lot more too.

Caster does gain but only in total if it's offsetting the KPI. It's not a formula but a guide is to start with 2 x KPI.


cymtriks - 21/4/10 at 07:06 PM

quote:
Originally posted by cheapracer
For a road car that doesn't run on silky smooth race tracks and needs suspension travel I would suggest a 2 to 4 degrees KPI , double that in caster, negative scrub of around 20mm +- 10mm and about 30 - 40mm of trail.

neg scrub is pretty easy to get these days with most wheels being for front wheel drive cars with plenty of offset that way.


In order to fit a disc and caliper into the wheel you are effectively forced to have a high KPI of around 12 degrees if you also have a small scrub radius.

Going to a negative scrub would push the KPI up even further given that the lower ball joint still has to be clear of the inner surface of the brake disc.

An overlooked reason to have castor on a road car is that it tends to unwind under braking. You need 4 or 5 degrees even with hard bushes to make sure that some is left under extreme braking.

So it looks as if the closest you could get to your ideal would be more like 15 degrees KPI, 4 degrees castor and 30-40mm trail.

Why do you opt for negative scrub? Positive is usually considered the choice for a sportscar (though this may be based on not a lot).


cheapracer - 22/4/10 at 05:56 AM

quote:
Originally posted by cymtriks

In order to fit a disc and caliper into the wheel you are effectively forced to have a high KPI of around 12 degrees if you also have a small scrub radius.

Going to a negative scrub would push the KPI up even further given that the lower ball joint still has to be clear of the inner surface of the brake disc.

An overlooked reason to have castor on a road car is that it tends to unwind under braking. You need 4 or 5 degrees even with hard bushes to make sure that some is left under extreme braking.

So it looks as if the closest you could get to your ideal would be more like 15 degrees KPI, 4 degrees castor and 30-40mm trail.

Why do you opt for negative scrub? Positive is usually considered the choice for a sportscar (though this may be based on not a lot).


Firstly you can not quote KPI angles without knowing the spacing between the BJ's - I figure you read that 12 degree figure somewhere in some article. A BJ spacing of 100mm will have a vastly different KPI angle than a BJ spacing of 200mm with the same vertical offset won't it. - I just looked, starting with a 12 degree example at 100mm BJ spacing, doubling the separation to 200mm gave me 7 degrees.

I haven't had any problem getting around my ideal with various uprights I have made up over the years but I'm also not limited by items such as stone shields, stamped control arms, oversized BJ's etc.. and for years now a wide range of wheels can take care of scrub if it's an issue for you. IF.

Look at it another way, cut 2 uprights in half horizontally and join the 2 halfs with the lower ball joints together and what have you got? Here is a nothing special Miata example I cropped, it would give 0 kpi, 80mm face offset with a fairly standard 40mm offset wheel gives some their ideal 40mm pos scrub. If I add maybe 5 degrees of kpi I will have some neg scrub.

Neg scrub is far more stable in every condition especially braking, shorter lever arm in deflection and the wheel steers in the opposite direction to the deflection force helping to cancel it. I'm not sure where your sports cars have pos scrub originates from in modern times but yes it can be a packaging issue sometimes.

Just a warning, don't go less than 10mm scrub either way, crossed signs are a bad thing.

A loss of caster under severe braking isn't an issue, the tyre has well enough saturation from weight transfer to be not interested in a couple of degrees loss of camber - why a you turning into a corner under that circumstance anyway .........


[Edited on 22/4/10 by cheapracer] Rescued attachment doublehalves.jpg
Rescued attachment doublehalves.jpg


Neville Jones - 22/4/10 at 07:24 PM

Top flight LM cars go for 0 everything, then dial in 'feel' as the drivers require, using castor. At least, that's how they were doing it a couple of weeks ago.

Scrub, is an evil. Wrecks the drivers arms with unnecessary movements and vibrations, and gives constant and unnecessary movements, sometimes confusing and covering up real issues.

Trail is used on supermarket trolleys, and belongs there only. Someone said somewhere that some LM car had zero everything, but used up to 65mm trail......??? Someone may also have been having his chain pulled by whoever made the initial statement. In motor racing, if you haven't seen it and measured it with your own hands and eyes, it's bovine excrement!! Motorsport is awash with the stuff. When repeated often enough, it becomes absolute truth, and everyone goes on their merry way.

KPI is optimised for centrepoint steering.( before castor is added). If you are lucky enough to be using wide enough wheels, then kpi can be as little as zero. KPI on fwd cars with Macpherson struts is usually due solely to packaging issues, and getting acceptable clearances.

Sometimes, kpi and castor can be utilised and optimised for cornering grip, or so I'm told.

Cheers,
Nev.


cheapracer - 23/4/10 at 05:35 PM

quote:
Originally posted by Neville Jones
Top flight LM cars go for 0 everything,

Trail is used on supermarket trolleys, and belongs there only. Someone said somewhere that some LM car had zero everything, but used up to 65mm trail......??? Someone may also have been having his chain pulled by whoever made the initial statement. .


The LMP2 winning Strakka HPD uses 0 KPI, 65mm of trail and minute caster and camber to suit each track, I know a development engineer who advises to the team, I chat with him often but thanks for your input. If you find a picture of it with the bodywork off you can actually see this large offset.

The last race car I developed for a client ran 45 mm of trail on the uprights I made, best thing I ever did and my current build is over 50mm of trail.

Important to have low'ish camber thrust and a bit of KPI for shimmy damping with high trail though.


cymtriks - 23/4/10 at 10:17 PM

quote:
Originally posted by cheapracer
quote:
Originally posted by cymtriks

In order to fit a disc and caliper into the wheel you are effectively forced to have a high KPI of around 12 degrees if you also have a small scrub radius.

Why do you opt for negative scrub? Positive is usually considered the choice for a sportscar (though this may be based on not a lot).


Firstly you can not quote KPI angles without knowing the spacing between the BJ's - I figure you read that 12 degree figure somewhere in some article.


No, it isn't something from a book. It is simply what you end up with when you have allowed for everything else.

Take a 205/50-15 size tyre on a wheel. Lets call it 8 inches wide.

The wheel spokes will take up the first 1.5 inches, the caliper needs some clearance say 0.5 inches and the caliper to the disc centre is at least another 1.5 inches so we now have the disc in the middle of the wheel. The lower ball joint must be inboard of that plus clearance for heat from the disc, say 1.5 inches inboard of the wheel centre.

As you correctly point out the lower pivot can be at any reasonable height but that still restricts us to something in the range 6 to 9.5 inches from the road surface.

To give zero scrub that forces a KPI of between 9.0 to 14.0 degrees.

Oddly enough the above description fits an Elise, with 12 degrees KPI and 10.5mm scrub fairly well. The McLaren F1 had wider wheels which allowed bigger offsets. It used 5/8 inch scrub and 9.0 degrees KPI.

The following is copied from the USA Locost site:

================... KPI ......... caster

Lotus Elise...................... 12° ......... 3.8° caster
Mazda Miata................... 11.3° ........ 5° caster
Ferrari F355.................... 13.16°
McLaren F1...................... 9° ........... 6° caster ...... 16.25mm scrub

Can you explain how you get around the packaging restraints that I've mentioned?


cheapracer - 24/4/10 at 02:48 AM

quote:


Can you explain how you get around the packaging restraints that I've mentioned?


I already gave you an example above with a sectioned Miata upright and thats a very limited starting base - huge castings, huge BJ, LBJ very close to axle centerline and a very wide disc offset.

I'm sure you can use your imagination from that info.


Doug68 - 25/4/10 at 08:03 AM

The attached section shows a cut through my C5 Corvette model. If my math is correct that 8.3 deg KPI and 12mm positive scrub.
With a bigger tyre the scrub could be reduced further but I think a 245 on the front is enough already
Of course this is all static, this is all bound to change on the road with tyre flex, camber etc.

Also I see caster removal on braking as a good thing, anyone whos riden a motorbike will know its a good thing to aid turning in.
Frankly though I think the ammount of change that happens in a car is insignificant.

Back on subject surely there's got to be parts in the world that make it not worth the effort to make your own uprights? From my point of view I've got enough bloody work to do on the car without adding on parts just for the fun of it? Maybe I'm missing the point though. Rescued attachment FrontUprightSectionII.jpg
Rescued attachment FrontUprightSectionII.jpg


Neville Jones - 25/4/10 at 10:15 AM

quote:

The LMP2 winning Strakka HPD uses 0 KPI, 65mm of trail and minute caster and camber to suit each track, I know a development engineer who advises to the team, I chat with him often but thanks for your input. If you find a picture of it with the bodywork off you can actually see this large offset.

The last race car I developed for a client ran 45 mm of trail on the uprights I made, best thing I ever did and my current build is over 50mm of trail.

Important to have low'ish camber thrust and a bit of KPI for shimmy damping with high trail though.


Pics of a corner with a wheel off, please, with a reference marker/graduated background alongside to show these things clearly.

Otherwise, just so much more garden fertiliser.

Cheers,
Nev.


cheapracer - 30/4/10 at 04:28 PM

quote:
Originally posted by Neville Jones


Pics of a corner with a wheel off, please, with a reference marker/graduated background alongside to show these things clearly.

Otherwise, just so much more garden fertiliser.

Cheers,
Nev.


I've given you the manufacturer's name and model and where my info comes from first hand so how about you show my information is false since your the one saying it is?

If it pleases you your Majesty I'll post pics of my own simple 50+mm trail upright's as I fab them in the coming weeks.


Neville Jones - 30/4/10 at 11:13 PM

Pictures of your own uprights are not pics of those you claim as running on the 'top' car.

It's quite simple; if you can't back up what you are saying with clear and concise pics of the claimed articles, then don't make the statement.

There's just far too much BS touted about as truth, and until the clear and concise pics of this LM car are put in front of me, the claims are just hearsay. Paddock BS to the agricultural among us.

I am close enough to a guy who designs this stuff, in front of me, to know when 'misinformation' is being put about.

Cheers,
Nev.


cheapracer - 1/5/10 at 12:37 AM

So your Mate is the designer of the Strakka?

Whatever. Ok, as this is obviously going nowhere lets take it from a different angle....

As I have already stated why trail is good you can counter using your depth of knowledge on the subject - simply complete these statements...

"Less trail is better because.....

or

"Too much trail is bad because.....


Doug68 - 1/5/10 at 05:45 AM

Handbags at twenty paces...


cheapracer - 2/5/10 at 08:10 AM

I don't need handbags, I can slap and pull hair all by myself.


Neville Jones - 2/5/10 at 01:39 PM

quote:
Originally posted by Neville Jones
It's quite simple; if you can't back up what you are saying with clear and concise pics of the claimed articles, then don't make the statement.



It's not for me to try to convince the world of my own opinions and experiences. Given that up years ago.

It's up to you to show proof of your statements on revolutionary geometry. Pictures of this Strakka upright would stop any alternative statements dead. You made the statement, you show the proof.

I could say that drunken Martians live in a USA built commune on the far side of the moon. Without solid pictorial proof, who'd believe it? Then there's always Photoshop....

And, the same effect as your 'trail' can be got from castor, with a few advantages ( or so I'm told) to boot.

I'm off to the next shuttle, to go have a beer with my Martian mates.

Cheers,
Nev.

[Edited on 2/5/10 by Neville Jones]


cheapracer - 6/5/10 at 04:48 AM

Neville Jones said: "Trail is used on supermarket trolleys, and belongs there only"

Listen buddy, you were the one that strolled into this thread saying that trail belongs on shopping trolleys - you have given no reason at all why you made this statement. Instead even though I state technical reasons why and give you Team names who do it yet you insist on calling out bullshit with no foundation.

Now you go on to say that theres other ways of getting trail - hang on you said trail was bad so why would you want other ways to get it? - well, is trail bad or not and WHY is it bad??

This thread is about building your own uprights, I'm trying to provide some helpful information for people to decide on their own designs - wheres your technical input?

I'll ask you again - Why is trail bad? The people considering designing their own uprights would be thankful to hear from you.....

My mate will try to take a snap next meeting he's there with Strakka but in the meantime you will have to do with an Australian V8 Supercars upright with around 50mm of trail picture (depending on caster of course).





[Edited on 6/5/10 by cheapracer] Rescued attachment upright2.jpg
Rescued attachment upright2.jpg


v8kid - 6/5/10 at 03:27 PM

Cheapracer how does that upright relare to the car? Is the caliper in front of the front axle or behind it?

I ask because if the calipers are behind the front axle, which is almost the convention here, there is quite a big offset which appears to give zero trail the way the photo is taken.

Either way is that level of offset between the king pin axis and the wheel center normal? It appears to be about 50mm and I assumed for some reason (that I can't for the life of me recollect) that it would be around 20 to 25mm.

Cheers

Davi


Neville Jones - 7/5/10 at 09:54 AM

Please, show a pic of one of those things ON A CAR!

How about a different angle pic of that one above, showing the opposite side of the brake disc, as well as the upright?

I happen to be quite close to the V8 community, and know the rules governing the parts used. You're playing in my backyard here, well and truly. If you understand what that means?

How long is the steering arm on that upright above? WHERE is the steering arm on that thing above?

From that pic also, that caliper doesn't belong on that upright. The disc looks as though it passes through the middle of the pistons. Maybe it's the disc that's wrong?

Is that a six piston caliper? Check a proper V8 for what is currently used.

And you expect everyone here to believe that a 3000lb race car that bounces over curbs, has it's front stub axle held into the upright with that single bolt?

[ with around 50mm of trail picture (depending on caster of course).]

Do you understand what you are talking about? Truly, really? Trail is independant of castor, is built in, and is constant. Centrepoint(or pressurepoint) offset, due to castor, is variable and is a function of castor. This is why castor is preferable to trail. Trail also can get 'the wobbles', just like a trolley. Castor, with the geometry and the method in which it loads, is self damping in this respect.



Good try, but no cigar this time.

That pic above is one of your own uprights, is it not?







[Edited on 8/5/10 by Neville Jones]


Theshed - 9/5/10 at 07:14 AM

Rather depends on how you use the word "trail" does it not? If you say trail is the measurement between the wheel centreline at ground level and the point at which a line through the ball joints intersects the ground then it is clear that trail will vary with castor. If you define trail only by reference to the upright i.e. the relative positions of the balljoints and axles then it cannot vary short of moving the balljoints.

I must say that the first definition is the one used in many suspension design programs.

Handbags down and see if you are squabbling about nothing much.


cheapracer - 19/5/10 at 10:37 AM

That is an older V8 Supercar upright but it runs basically the same geometry then as one does today as the tyres have remained fairly constant for a few years. Todays brake package is different as they have a regulation brake now they must use.

The caliper is at the back of the upright and the stub axle is to the front because Supercars run in the order of 12 to 13 degrees caster the trail ends in the region of 50mm varying slightly with caster tuning.

They also run about 6 degrees of Camber.


cheapracer - 19/5/10 at 12:07 PM

quote:
Originally posted by Neville Jones
Please, show a pic of one of those things ON A CAR!

[Edited on 8/5/10 by Neville Jones]


Heres a BF Falcon V8 Supercar upright, brakes etc on a current build Sports Sedan.

For those who are truly interested notice because of the caster used and the axle ends up aprox directly above LCA, you can see just how much trail it has by the time the caster pivot line hits the ground.

Whats next Mr Jones - video? stat declarations?

Why is trail bad Neville? You still haven't answered the question....... Rescued attachment BF.jpg
Rescued attachment BF.jpg


Neville Jones - 19/5/10 at 06:19 PM

Why is trail bad Neville? You still haven't answered the question.......

I think if you go back a few, my answer is there. If you can read and understand plain english.

If you can see all that, in that grainy pic above, then you should be working for MI5, interpreting satellite pics of our enemies.

That pic shows a remarkably normal upright,(as far as a man can interpret it) with remarkably normal geometry, if you look closely. No 'trolly trail' on that item.

I'll have to take your word that the photo is what you reckon it is, though, but I have my doubts.

I was in Brisbane a few months ago, and got a tour of the workshop of a top team. I'll be back there in a few weeks, and living but a short walk from that same team. I'll have to pop in and see if they've made any dramatic changes in the last few months.


Cheers,
Nev.







[Edited on 19/5/10 by Neville Jones]