Talk:Understeer and oversteer

Gradients
A gradient, often written as dX/dY, measures the ratio between the variation of a quantity, in this case X, and the simultaneous variation of another related quantity, in this case Y. All the gradients the article is talking about have an angle as X and the amount of lateral acceleration as Y. An expression like "based on the gradient of the steering needed to make a turn in a steady-state condition" is not just imprecise, it makes simply no sense, because it states just what the X quantity is and not what the Y quantity is. It is like saying that two is "the difference between five", it makes no sense unless you add "and three". Moreover the word "gradient" although technically correct (when used properly, i.e. not as in the article) is confusing or outright obscure to most readers (just pretend not to know what a gradient is and try to read the wikipedia article). I would suggest a wording like:"based on the amount of correction to apply to the steering angle needed to compensate for a given lateral acceleration". Any definition, by the way, is better than one that makes no sense, so now I will make a minimal correction. Could please an expert review this and update the article?

Also the text is rather obscure about the different gradients involved and fails to name the Y quantity (had the author a decent understanding of basic calculus?). The only mention of the unit of measure of U is en passant in a negative phrase:"Results depend on the type of test, so just giving a deg/g value is not sufficient; it is also necessary to indicate the type of procedure used to measure the gradient." What the article does in the section "Understeer values" contradicts this very phrase: no mention of the procedure. Moreover, for what I can understand, the values given in that table are degrees of rotation of the steering wheel per G, not consistent with the definition given in previous sections.

— Preceding unsigned comment added by 89.153.57.129 (talk) 15:02, 13 September 2011 (UTC)


 * You're right, the only place a gradient is defined it's wrong (a ratio). I wrote that section, copying text from SAE, ISO, and Gillespie and got it garbled. It should be technically correct now, even if fairly technical.


 * I'm not able to maintain this article or contribute much more, but here are two observations:
 * 1. You're right, the table of values are not consistent with the definitions. They should be removed.
 * 2. Also, the oversteer figure is very misleading. When a trajectory like the one shows occurs (a tighter radius), it's because the driver slowed down a lot during a loss of control. If the speed is not changed, the trajectory of the vehicle for an oversteer failure to follow the target radius is about the same as an understeer failure. The difference is that the vehicle is facing forward in understeer and is spinning out in oversteer. — Preceding unsigned comment added by Mikesayersskier (talk • contribs) 04:57, 25 October 2011 (UTC)


 * Mikesayersskier (talk) 04:59, 25 October 2011 (UTC)


 * Hello Mike. As for the figure: I am the one who added those images. I spotted them at the Dutch Wikipedia (here and here) after trying to discover the differences between the two terms. I found the pictures very helpful in understanding it, and I was surprised that they weren't used on the English Wikipedia. If your analysis is correct, however, I understand why. Forgive me my utter ignorance on this subject, but is the oversteer figure truly "misleading", as you say? Cheers, theFace 18:29, 27 October 2011 (UTC)


 * Disagree. The diagrams are NOT showing the effect of a loss of control. They show the affected trajectory with uncorrected understeer and oversteer. After a loss of adhesion there is no steer of any kind. -- Ian Dalziel (talk) 21:22, 27 October 2011 (UTC)


 * The racing drivers joke is correct, in practice. You normally go through the outer fence, the question is which bit of the car hits first. The diagram shows a case that is hard to explain, since the lateral acceleration seems to be increasing despite a loss of grip, for oversteer. OK it isn't that simple and that trajectory is possible, but it is certanly not something I've experienced in real life.Greglocock (talk) 21:56, 27 October 2011 (UTC)


 * The only reason you don't experience it in real life is that you correct it by moving the wheel. It's just oversteer before any loss of grip. -- Ian Dalziel (talk) 22:04, 27 October 2011 (UTC)


 * Looking at the current page, the captions are wrong. The diagram doesn't show "a spin", just the normal oversteer effect. the spin happens when an oversteering car exceeds the limit of adhesion. -- Ian Dalziel (talk) 22:07, 27 October 2011 (UTC)


 * As I said, I'm ignorant on this matter and didn't knew better. It looked like a spin so that's what I called it. ;-) Feel free to change the captions. - theFace 15:27, 28 October 2011 (UTC)


 * Both diagrams are at odds with the engineering definitions of understeer and oversteer (ISO and SAE), because the engineering definitions apply only while the vehicle is under control while following a circular path at nearly constant speed. There are different tests used to characterize the understeer gradient, but if the diagram shows a road then the assumed test is one where the radius is fixed and the speed is increased slowly to get gradually increasing lateral acceleration. The understeer/oversteer comes from how much you have to steer to follow the path. Once the vehicle diverges from the path, the test is over. I've seen the diagrams in other places, mainly in marketing for stability control systems. They are probably tied to the way understeer and oversteer are used less formally to describe how a loss of control occurred. But unless you slow down, physics won't let you tighten the turn as in the diagram for oversteer, which is why I think it's misleading. (I guess I'm also saying both are misleading, because the formal definition of understeer gradient doesn't involve trajectory.)


 * BTW, a few months ago I put up some figures showing simulated vehicles at understeer and oversteer limits. They got yanked by Wiki police for possible copyright issues by someone unfamiliar with the type of software I used (CarSim and its animator tool). Oh well. Mikesayersskier (talk) 16:41, 29 October 2011 (UTC)

Complete Replacement
Understeer and oversteer have been well defined by automative engineers since the 1960s at a primary descriptor of steady-state cornering. Although the terms are also used by enthusiasts of cars and motorsports to apply to other maneuvers, they usage should be consistent.

The existing article is wrong from the 1st sentence. (front/rear tire slip angle relationships are determined more by C.G. location. If the mass is forward the front tire need more side force and hence more slip.) The descriptions that follow have more to do with limit behavior than general steering response.

The article I just put in is much shorter and perhaps more technical, but it's not wrong. If it survives the Wikipedia environment for a few days, we can add more content (plots of understeer, animations of understeer/oversteer comparisons, etc.).

If you are reading this and are interested in the subjects, please refer to the three references cited: J670, ISO 8855, Gillespie textbook.

I didn't make this stuff up; the terminology and these definitions go back at least to 1976, as covered in the previous J670 standard. Mikesayersskier (talk) —Preceding undated comment added 21:56, 28 December 2010 (UTC).


 * Well, it is a bit rude to completely rewrite an article with no prior warning, but yes, the previous article was pretty bad. Greglocock (talk) 22:00, 28 December 2010 (UTC)

Sorry, I'm pretty clueless about Wiki conventions. It's too late for permission so I guess I need to look for forgiveness.

Follow Up
I know it's drastic and possibly a little rude to completely replace an article, but in this case the majority of the text was misleading or outright wrong. Looking at the version from Dec 27:
 * 1) The first sentence is completely wrong, as noted above.
 * 2) The picture is wrong. It shows a limit failure where the understeer gradient is infinite (the car is incapable of reaching the lateral acceleration for determining the gradient).
 * 3) The explanations of what cause understeer are partially correct, but miss the main factors (tires, compliance, load transfer, kinematical steer).
 * 4) The Physics section is wrong, again stating that understeer is based on front/rear slip. It's not.
 * 5) The section on limit handling is using understeer to describe highly transient maneuvers. This is at odds with the official definition used by automotive engineers, but does represent usage by race teams and non-professional enthusiasts. It could be put back in (heavily modified) after introducing the alternative usage of understeer and oversteer to describe a general tendency of a vehicle to turn more or less than expected. For example, applying power in a racecar with locked differential in a turn puts the power on the inside rear wheel, forcing the vehicle out of the turn. This effect is considered understeer by race teams. Similarly, removing throttle puts engine braking on the outside rear tire, also forcing the vehicle out of the turn for another understeer effect. These are fairly advanced details, and I would normally not put them in a basic article where someone would go to get a simple explanation of what understeer and oversteer mean.
 * 6) The table of values is providing details without a valid explanation of what they mean, or how they were measured. All they do with the article as it was (or as it is now) is add length.

Some of the discussion implies that potential authors are trying to learn very basic concepts of dynamics and vehicle dynamics in particular. This is fine, but I would think that encyclopedia articles should be written by experts who are already versed in the subject. If you're not an expert, read and comment, but don't write other than to correct typos and styles.

The article in its current state (Dec 29) is pretty minimal. If it is to be extended, it should first have examples of how understeer gradient is used, with plots such as those in J670 or Gillespie's textbook.

As mentioned at the top, understeer and oversteer are the same measure: if the gradient U is positive, it's understeer; if negative, it's oversteer. This article is intended to cover both terms. The oversteer article is longer, and in general, has even more errors. I don't know how to merge them, but my recommendation is to remove the oversteer article and work with this one. Mikesayersskier (talk)


 * Two comments to begin with. The diagram doesn't show a limit failure, it shows an extrapolation - see the matching Oversteer diagram. Then "encyclopedia articles should be written by experts" - ABSOLUTELY NOT. See WP:OR. Articles should be fully sourced, not newly written, whether by experts or by me! -- Ian Dalziel (talk) 17:04, 29 December 2010 (UTC)

Copied from Oversteer (May 18, 2011):

FWIW, I still think this whole article should be merged with Understeer. Mikesayersskier (talk) 00:15, 18 May 2011 (UTC)


 * Do it already. Just follow the directions here. -AndrewDressel (talk) 02:12, 18 May 2011 (UTC)


 * OK. No one's objected in 5 months. I'm trying a redirect from the Oversteer page. Mikesayersskier (talk) 03:07, 18 May 2011 (UTC)


 * Nice job. -AndrewDressel (talk) 12:55, 18 May 2011 (UTC)


 * Great job on the combination of the two articles. It was a good idea. I have a couple of comments for you:


 * Change the article name. It doesn't have to be called "understeer and oversteer" but it might be called "vehicle balance" or something along those lines.
 * Explain the meaning of understeer/oversteer better without the use of technical terminology. You have a great section on the technical definition according to the SAE, but 99% of the people reading this article will not understand the terminology in it. The explanation for the average reader could be fleshed out. Additionally, link more of these technical terms so people can learn about them if they want. I will take a shot at this later when I have more time, if you want.
 * I think this article could be longer without hurting its comprehensibility.
 * Altogether, it was a good move, thanks for taking the time to fix it up. Bdc101 (talk) 14:47, 19 May 2011 (UTC)


 * Agreed, the merge has made it much better. However I disagree with you about the title, it should be the common search term for the concept. I don't think many people would search for vehicle balance, and if they did I think they'd be surprised to read about understeer. Greglocock (talk) 23:44, 19 May 2011 (UTC)
 * Well, that was just a suggestion, but what about the people who search for "oversteer" and get redirected to "understeer?" My suggestion is to make both terms redirect to this article, but not have the article named after only one of them. Bdc101 (talk) 14:54, 20 May 2011 (UTC)


 * Maybe redirect to a new title Understeer and Oversteer? Mikesayersskier (talk) 15:19, 20 May 2011 (UTC)


 * Nearly there. The idiot wikilawyers will insist on 'Understeer and oversteer or vice versa. Which is illiterate but self explanatory. Greglocock (talk) 02:25, 22 May 2011 (UTC)

Table
The section on actual understeering values says that "How much a car understeers can be measured in the number of degrees more the streering wheel have to be turned per G of force. Here are values for some cars.[1] The higher the number the more the car oversteers.". Howcome having to turn more rather than less means it oversteers? Shouldn't it read that "the higher the number the more the car understeers"? 62.169.74.64 03:45, 30 August 2006 (UTC)


 * Yes, you are right. I can't say I really like that list, there is no source given, and some of the number seem a little odd. I've measured a few of those cars, I'll check on Monday. Greglocock 06:17, 9 September 2006 (UTC)


 * Unless it has been removed there should be a source. The problem is that the web source don't give the full list. It's only available in the paper version. // Liftarn


 * how is this being measured. do all the cars have the same tyres or their standard tyres. because i don't believe, though could be wrong, that the Opel (vauxhall) Astra understeers less than a Impreza. if u want to compare the Astra VXR and the Impreza, which have similar power, im sure the astra understeers much, much more. let alone the torquesteer. Pratj 23:36, 9 September 2006 (UTC)


 * Usually you measure the linear range understeer by plotting the steering wheel angle (SWA) versus the lateral acceleration, whil driving around a constant radius, typically 30-61m. Usually you take the average gradient between 0.1 and 0.4 g, for a car. An alternative is to perform a swept steer, driving at constant speed and gradually increasing the SWA. These two methods give slightly different results. Many modern cars with good handling have a lot of understeer in order to improve their high speed performance, and yaw damping. These tests are done on the standard tires otherwise they are meaningless- changing tires can affect u/s gradient by 15 deg/g. Note that this a steady state linear range measurement, it does not addresss torque steer or limit handling. Greglocock 00:05, 10 September 2006 (UTC)


 * steering wheel angle has nothing to do with understeer/oversteer. My car has an 11:1 steering ratio, a friend's has just under 17:1. How could you possibly compare those to each other with that method? Furthermore, how does that relate in any way to how the tires are behaving?


 * Shrugs. Some people measure u/s at the steering wheel, others divide the SWA/g by the nominal steering ratio, others do other things. I was working with an approach consistent with the figures in this article. Greglocock (talk) 04:45, 26 December 2010 (UTC)


 * Incidentally, if you actually look up s real vehicle dynamics book, like Gillespie, you will find that you are wrong. 100% wrong. 00:55, 27 December 2010 (UTC)


 * As it says on http://www.automobil.se/zino.aspx?articleID=10863 They drive the car around a curve of 50 m radius. First at 10 km/h, then increasing the speed. First they measure how direct the steering is (i.e. how much you have to turn the wheel at low speed). Then they plot a curve (see bottom of page) of steering angle (vertical) and side acceleration (horizontal). The slope of the curve shows how much the car understeers. Toward the limit of the car the curve turns abruptly showing at what speed the car terminally understeers (or sometimes terminally oversteers). // Liftarn


 * i understand. but would i be correct in saying that this by no means is a definite clue as to how a car would understeer under race track conditions? Pratj 14:43, 10 September 2006 (UTC)


 * Both yes and no. The car may oversteer if it looses grip on the rear wheels (for instance if applying too much gas in a curve). The values do show how much the car understeeers, but it's not entierly linear (even if most cars are very linear, the Morgan bean an exception) so depending on the side acceleration it may both oversteer and understeer on various conditions. // Liftarn


 * i think a mention of whether the car is rear wheel drive or front wheel drive should be made. as some of the rear wheel drive cars as u say will have a tendancy to power oversteer, wheras the front wheel drive cars will tend to understeer under most circumastances. Pratj 17:35, 10 September 2006 (UTC)


 * Hang on, these numbers mean very little in the context of a racetrack. They only apply over the linear range (where the tires are linear) - typically up to say 50% of the maximum capability of the car. The linear range understeer tells you NOTHING about limit handling. Greglocock 02:52, 11 September 2006 (UTC)


 * Shall we mention that this is by no means a measure of limit handling then. as i, and im sure many other ppl, would be slightly mislead by those results. Pratj 14:15, 11 September 2006 (UTC)

More values
these are for discussion only, as I can't give a source for them.

http://www.edccorp.com/pdfs/WP2004-3.pdf#search=%22%20%22understeer%20gradient%22%22

for some more examples

As you can see the Swedish measurements seem to be way off base. There are good reasons to believe that the 350Z could not be so low. I don't know how to resolve this.

Greglocock 02:52, 11 September 2006 (UTC)

External Link
I've added an external link to a graph on my website. If it is inappropriate delete it. It is sort of original research, but more accurately a demonstration of a well known equation. Greglocock 05:33, 25 September 2006 (UTC)
 * Hi Greg - not sure if anyone without some serious technical knowledge would understand this one, think this section needs to be simplified where possible Driver sam (talk) 15:46, 26 May 2008 (UTC)
 * Aye, mediocrity will always defeat excellence. Greg Locock (talk) 04:18, 27 May 2008 (UTC)

Diagram
i moved someones diagram request from the main page to the talk page,

if we make a diagram how do we use it on wiki? Pratj 12:27, 14 October 2006 (UTC)

i made this image. its a bit basic but it illustrates what understeer is (if u can spend more time and make a better 1 please do, i only had 5 mins). i just don't know how u edit the page to include the picture. so if someone would show me or do it themselves i would be grateful. Pratj 20:25, 14 October 2006 (UTC)


 * I don't know how to illustrate understeer, as opposed to oversteer. Your diagram is as good as any. The difference, in an accurate diagram, would be a tiny change in angles. Anyway, to add an image, look in toolbox on the left under the search box. Upload your file to the server. Then when you are editing hit the "editing help" link in the row of buttons beneath the edit box. This tells you the command to link to an image. Well this is baffling, you've done it? I'll move it. Change the words if you like. Greglocock 20:44, 14 October 2006 (UTC)

why has the diagram been deleted de to copyright? i made it so surely it belongs to me? Pratj 13:48, 24 October 2006 (UTC)


 * Perhaps you didn't state it as clearly as you should have. // Liftarn


 * Yeah, thanks for that (sarcasm). Pratj, can you load your picture file up using the media links, and remember to click on the button that says that it is your own original work etc. Then reload it into the page as before. Greglocock 19:45, 24 October 2006 (UTC)


 * I don't think I have gotten anything deleted, but I have had peaople asking for source when it clearly says "Taken by myself". Some people don't seem to check things first. // Liftarn

can't u just revert it. if not i will load it up again. Pratj 21:55, 24 October 2006 (UTC)
 * The important thing is that the image has a clear copyright/license tag, such as GFDL-self or similar - as mentioned on the upload page. The automatic scripts and editors who check these things will delete an image without a copyright/license tag. -- Solipsist 10:00, 25 October 2006 (UTC)

i did that last time, but i'll do it again sometime today Pratj 10:24, 25 October 2006 (UTC)

Can someone please include a diagram which illustrates the phenomena clearly? 143.53.7.186 02:54, 7 December 2006 (UTC)

Removing references
Go on, tell me why the reference to Carroll Smith was removed? Most articles I write get pinged for not including references Greglocock 21:46, 27 December 2006 (UTC)

Tyre placement
Not very happy with the "Common Practice" section. Fitting new tyres to the front doesn't increase understeer, it usually reduces it. The potential safety problem is that it can induce lift-off oversteer. -- Ian Dalziel 16:25, 1 January 2007 (UTC)

yeah, i agree. the problem is understeer normally results in increased front tire wear. so u have no choice. Pratj 18:33, 1 January 2007 (UTC)

Pratj- I'm glad you are sorting that horrible para out, the one that Liftarn hit with the BF cite G. As to Ian's point The correct thing to do is to move the least worn old tires to the front axle, and fit the newies on the back. There's optimism for you Greglocock 23:06, 31 January 2007 (UTC)


 * Indeed - but that isn't an answer to my point, it's agreeing with it! Fitting new tyres to the front does NOT increase understeer, so the item is completely wrong, reference or no reference. The reference doesn't mention understeer, by the way. -- Ian Dalziel 06:39, 1 February 2007 (UTC)

Understeer values citation?
There's a citation in "Understeer values" but it doesn't link to anything. Disappointing, as I've like to know where the numbers came from...


 * They came from the sports car magazine Automobil, but the link will do no good since the list is only published in the paper version. // Liftarn

I concur - deleted the section. Somebody can bring it back if they have a real understanding of the understeer coefficient. The section as-is is pretty much nonsense; a quick google search turns up nothing when looking for "understeer value" and "understeer rating", and the maths for "understeer coefficient" are much more complicated than implied by the headline of the section. — Preceding unsigned comment added by Sniper1rfa (talk • contribs) 13:44, 24 December 2010 (UTC)

Dubious
"However, in certain cars and under certain circumstances(e.g when there is a small amount of grip still available) the understeer can be reduced by applying full power to the wheels rather than braking"

This either needs to be amplified up into an explanation of when and how, or deleted, in my opinion. At the very least is the editor talking about linear range or terminal? front wheel drive or rear? It just looks like an anecdote to me. Actually the whole of that Design section needs a complete rewrite, and a better structure. I'd suggest separating FWD and RWD, linear and terminal, etc etc. Greg Locock (talk) 22:57, 22 January 2008 (UTC)

It's just completely false, except for rear-steering cars. Understeer is always increased by acceleration or reduced braking, always decreased by braking or reduced acceleration.


 * I m talking about a front wheel drive car. And I m talking about when there is some grip left available. It sounds really wierd but if you ve wathced for example MotoGP (not entirely irrelevant) you may have seen cases where the bike seems to be going off track, and the driver applies full power and keeps the bike on the track. The same thing happens with my car, and possible many others. It is a Nissan Almera and I can kill the understeer by accelerating. This DOES NOT apply when the car has reached its physical limits of grip. Unfortunately I don't know how I can explain it better or give a citation. 147.27.47.82 (talk) 10:16, 23 January 2008 (UTC)


 * I think MotoGP and FWD are, to say the least, a hard stretch to combine. Shall I go and ask a real live MotoGP tire engineer? Is an Almera FWD or RWD? What you are describing is throttle-on tuck-in -  I see it all the time in a RWD on gravel, although not on the limit. Greg Locock (talk) 10:29, 23 January 2008 (UTC)


 * Take a look at this, maybe it will help http://www.iwsti.com/forums/driving-tips-techniques/22357-do-you-ever-accelerate-midst-turn.html The user WE_R_ND describes that when he accelarates in a corner his car doesnt push to the outside of a turn. It is about an awd car.
 * Ok maybe I havent understand well what the term understeer includes. (I'm definetely not talking about terminal understeer). Maybe the best thing to do is delete it...147.27.47.82 (talk) 10:48, 23 January 2008 (UTC)


 * Well, BBs are a lousy reference. What they talk about is real, but way beyond the scope of this article. that's a transient event. Greg Locock (talk) 11:02, 23 January 2008 (UTC)

wow, been a year since i left this article, i thought it was in good shape then. anyhow:

yes, the statement that actually applying full power can actually decrease understeer better than braking seems odd to me... braking can actually cause snap oversteer as the shift in weight to the front wheel provides better traction to the front wheels (and makes the back end light) causing a spin. i have actually heard of very few cases, particularly since the emergence of good ABS on modern cars, of braking actually increasing understeer but i do accept there is a limited possibility. however, applying FULL throttle is never going to reduce understeer. lets say i have 100 "grips" (G) from my front tires. with no throttle 100G are going towards turning the car, with some throttle 20G to transfering torque to the ground and 80 G for turning. accelerating hard around a corner it may be 50G torque, 50G turn... but as u see, from this basic formula i just invented that u only have a limited amount of grip, and the more grip going towards "moving" the car the less there is for turning resulting in understeer.

however, all of this is based on a front wheel drive car. it is possible to reduce understeer by deliberately spinning the rear wheels on a RWD car. therefore all you are doing is disguising understeer with power oversteer.

I think we should remove the statement, or make it a lot clearer unless some form of evidence can be found for it. Pratj (talk) 19:16, 28 January 2008 (UTC)
 * It's not just a FWD/RWD issue, but also the engine orientation that affects behavior in the corner. If you're driving a mid-engine car, then letting up on the gas or braking during a hard turn can result in snap-oversteer as mentioned in the previous sentence.  In this type of vehicle, understeer may be what tempts you to slow down.  The risk of snap-oversteer is the reason not to. 24.19.238.74 (talk) 09:07, 9 April 2008 (UTC)

It's just wrong. Slowing down is not bad, you fools, it reduces understeer. AS ALWAYS, gradual changes are required. Simply taking your foot completely off the gas is wrong, pull gently away until you get steering. Also, of course, reduce steering input at the same time (i.e. steer into the understeer) so as to regain control. Yes, spinning the rear tires is an arguable substitute for shifting weight to the front, but you should say "spinning the rear tires" and you should also turn harder so as to increase friction at the front end, and oh yes, you better actually have adequate lateral inertia so the rear end skitters out the way you want it.

It's quite obvious from the defense above that this is his original research based on analyzing youtube videos and interviews and his almera. Toss it.

Oh shit, there's one more counterexample: when speeding up gets more downforce from a front spoiler (proportional to any rear spoiler). Hmm, more special cases: Pulsing the acceleration will rock the car back and forward, and the generally that will have a net steering effect. But it's not smart. In a front wheel drive car, more power to the wheels could have any of a number of effects on the suspension geometry (as the wheel pulls forward relative to the frame) which might include camber changes which could give it more grip. Fixing the suspensions is a better option. Oh yes, if genius boy is braking into the corner and his braking system is putting more pressure front that rear then he may be losing front end grip first and understeering; letting off the brakes might be giving him his traction back.

First, we should keep the serious definition of "under/oversteer" in this discussion--only caused by the ratio of front/rear tire slip angle. Following this definition, weight transfer is the only thing affecting under/oversteer. While there is no wheel spinning allowed, acceleration will move the weight backward on the rear wheels, increasing rear wheel slip angle, decrease front wheel slip angle, that means decreasing understeer or increasing oversteer. For FWD, acceleration may not make significant change for low limit of wheel spinning, but RWD can do more. Second, we are talking about how to save your life when entering a corner with unexpected pushout (not understeer; understeer happens only when tire grips) happens, is acceleration helps? This is no more an understeer question for what happen next is not gently accelerate, but floor the throttle. The main propose is to make the drive wheels spinning toward inside of the corner, making more force than tire grip when conering. It can be calculated in simple physics that in some given conditions, spinning wheels accelerate faster than grip by calculating effective power output (on wheel/road interface). These conditions includes: the max. static friction is small, the dynamic friction is smaller than static for sure, but close to it; wheel spinning RPM capability is high; drivetrain does not overheat. The most possible case I thought is on ice sufface. —Preceding unsigned comment added by ChengPingYu (talk • contribs) 16:21, 13 November 2009 (UTC)


 * "Following this definition, weight transfer is the only thing affecting under/oversteer." Wrong. read Bundorf's paper or milliken. US=weight transfer, tires, various compliances and kinematic effects. Simple example- put grippy tires on front, oversteer. Put grippy tires on back, understeer. No weight transfer changes at all. Greglocock (talk) 00:32, 14 November 2009 (UTC)

How can you change tire while you're driving that car? When we talk about this issue, the car is under an in-too-fast situation, and you want to change tire at this moment? —Preceding unsigned comment added by ChengPingYu (talk • contribs) 06:50, 15 November 2009 (UTC)


 * I'm talking about the engineering definition of understeer. It is an intrinsic property of the car. Greglocock (talk) 07:20, 15 November 2009 (UTC)

I think it should be removed the article is not meant to talk about advanced driving techniques that may or may not affect under steering, if people feel this technique is worth writing about then they can create a separate article for it. Different cars react differently in different situations anyway so what works with a one car may not work with an another, with no context such a claim cannot be verified true or false anyway. Very little gravitas indeed (talk) 22:56, 24 December 2009 (UTC)

Possible links?
Hi - I was reading this article and it seems to have quite an engineering focussed style of definition. I've done a but of research and found a few external references which may be useful for the reader with less of a physics / maths based background. There doesn't seem to be much out there but:

http://news.bbc.co.uk/sport1/hi/motorsport/formula_one/car_behaviour/default.stm - a BBC explanation of understeer and oversteer

http://www.drivingfast.net/car_control/understeer.htm - a dedicated article on understeer

Driver sam (talk) 11:57, 26 May 2008 (UTC)


 * Well I guess it depends on whether you want a schoolkid's encyclopaedia or a more generally useful one. The first article is too simplistic by far, the second one is confusing and unhelpful (it could be summarised as find what you are doing that causes the underster and then stop doing it). But it does at least suggest that things are not straightforward. Greg Locock (talk) 04:15, 27 May 2008 (UTC)


 * Many people use wikipedia as a first port of call when looking up automotive terms such as understeer and oversteer (probably including schoolkids!). Such terms, although commonly used in engineering are also useful for driving enthusiasts and anyone interested in motorsport. I'm sure you have an engineering background, but launching straight into relatively complex terminology doesn't help most people unless they have a residual knowledge on these topics. I'd welcome your suggestions on an external link which might help provide a more basic explanation, or I think it would be beneficial to have a 'dummies guide' to understeer at the start, same for the oversteer section. I'm not suggesting rewriting the article, but simply adding some more beginner friendly info. What did you find confusing in the second link, do you think there anything which could be re-written for this section? Would be interested to see what others think, or am I way off the mark here? Driver sam (talk) 16:17, 27 May 2008 (UTC)


 * I'm going to add the latter of the two articles primarily due to the decent illustrations. But as Greglocock has suggested, I do think there may be better links around so please feel free to suggest some. May do the same for oversteer, but would be great if someone could come up with a decent image for wikipedia on these topics. Driver sam (talk) 16:01, 10 June 2008 (UTC)

This statement is simply wrong if you exclude possible punctures, tire deflations etc (even then it is extremely unlikely). Braking will result in a weight transfer towards the from wheels which will allow them to generate more lateral force. In other words, braking will allow the front wheels to grip the road more. Although it is necessary to say that this is only valid where the slip angles are slightly higher than the peak lateral force slip angle (about 6 to 7 degrees). This also means that understeer can be reduced by reducing the steering input (provided the peak lateral force generation slip angle is exceeded in the first place). —Preceding unsigned comment added by 87.194.114.209 (talk) 23:11, 7 May 2009 (UTC)

Racing Understeer
In racing there are several sorts of understeer: straight line linear range understeer, understeer on entry to the corner (under trail braking), understeer at the apex, and understeer on exiting the corner, due to throttle. Until you know which one you are talking about then things will be very confusing. Greglocock (talk) 02:54, 8 May 2009 (UTC)

Understeer and Oversteer should be one topic
The concepts of understeer and oversteer have been around for many decades, and they're much simpler in concept than the two wiki pages indicate. The US Society of Automotive Engineers (SAE) has published a standard for terminology J670 that covers the topics in a few sentences. The definition in J670 from 1976 has been kept in the updated version (2008?) and is also in the new ISO 8855 standard on vehicle dynamics terminology. They're also in any textbook I've seen covering vehicle dynamics. Certainly Gillespie's "Fundamentals of Vehicle Dynamics" textbook (referenced in the Oversteer page) presents the concept succinctly.

Summary of the Definitions
The concept is that a car making a turn with non-slipping tires has front-wheel steer equal to the arctangent of the quotient of wheelbase divided by the turn radius. This is called the Ackermann steer angle. If the lateral acceleration increases due to increased speed on a constant-radius track, or decreased radius at a constant speed, or any combination, there is a gradient of steering wheel angle vs. g's that would be obtained with Ackermann steering (Ackermann's name is always used for the reference neutral steering).

The real measure is the difference in gradient (real - Ackermann). If the difference is greater than zero, it's understeer. If less than zero, it's oversteer.

Different companies and organizations have slightly different test procedures for establishing the gradient. It's normally done under quasi-static conditions in which either steer or speed are increased very slowly. Results do depend on the test, so just giving a deg/g number is not sufficient; you also have to indicate the type of test used.

In more transient maneuvers, such as racing turns or stability control tests, the names understeer and oversteer are also used by engineers to indicate that the steering needed for the maneuver is more (understeer) or less (oversteer) than some reference.

What Next
I'm new to wikipedia editing, and have no clue about how to merge pages. To make sure this is seen, I'm pasting it into the talk pages for both topics.

If they are merged and redone, I can contribute if needed to ensure that the descriptions match the SAE and ISO definitions.

Mikesayersskier (talk) 18:34, 12 June 2010 (UTC)

Merged with Oversteer
Article merged with Oversteer: See old oversteer talk-page here — Preceding unsigned comment added by Mikesayersskier (talk • contribs) 03:08, 18 May 2011 (UTC)

Comments on horsepower and torque
I have moved these comments here because the first time I removed them, they were simply reinserted.
 * "An old bit of racing humor says that an understeering car goes through the fence nose first, an oversteering car goes through the fence tail first, and with a neutral-steering car, both ends go through the fence at the same time. In the same vein, horsepower is defined as how fast the fence is hit and torque as how far the fence is taken by the car."

The fact remains that they are not about understeer, oversteer, or steering at all and so have no place in this article. If I am mistaken, please explain how. -AndrewDressel (talk) 18:52, 18 July 2010 (UTC)

This is the ONLY part of the entire effin article that makes any sense whatsoever and isn't a cringeworthy, obstructive, and uninformative obstacle course for both reader and eyes.208.127.80.59 (talk) 06:34, 7 September 2011 (UTC)

Move to "Understeer and oversteer"
I have renamed the article. I wasn't sure about the title, but eventually settled on this after doing a google test: "oversteer and understeer" gives 231,000 hits, "understeer and oversteer" gives 364,000 hits. Cheers, theFace 14:27, 25 September 2011 (UTC)

Understeer and Oversteer at the same time
How does one classify the following situation?

A car approaches a corner at a rate of speed that, even with max braking, cannot slow the car down enough to clear the corner without understeer. During braking, one of the rear wheels gets too light and starts to lose traction, putting the car in a slide. The car misses the corner and plows outwards, albeit with its rear end coming around at the same time. — Preceding unsigned comment added by 174.47.162.2 (talk) 19:08, 28 April 2014 (UTC)


 * I think your description is a good summary in itself. To be honest if the driver keeps the brakes on, or uses the handbrake, to try and get the rear sliding and bring the nose round then that is a perfectly valid technique, good fun in the snow and on gravel. Understeer is when the front slip angle is greater than the rear, or the yaw velocity is less than it should be for a given lat acc, oversteer the opposite, you have more or less made the car more neutralGreglocock (talk) 01:24, 29 April 2014 (UTC)

Two spellings of "Ackerman(n)"
I know nothing on this topic, thus my looking it up. I did note one spelling inconsistency between "Ackermann steer", found in the section titled Vehicle dynamics terminology and "Ackerman angle" in the section titled Related measures.

Seeing now that the link goes to a page called "Ackermann steering geometry," it is very likely the link's spelling of the family name is correct, but I will still let someone with more knowledge on the topic correct the spelling inconsistency, for it could go either way. Lytzf (talk) 00:07, 29 April 2015 (UTC)


 * I'll vote for nn see http://www.wolfgang-pfaller.de/landespa.htm  Greglocock (talk) 01:47, 29 April 2015 (UTC)