Why You Can’t Judge How Dark it is on Television

 Light, Optics, Talladega, Weather  2 Responses »
May 062013
 

When you were a kid, perhaps you locked yourself in the bathroom, turned out the lights, positioned yourself in front of the mirror and then turned on the lights to watch your pupils grow.  And if you’ve never done this, shame on you for not being curious.  Go do it.  Now.  Or maybe if you’re not the participatory type, you’ve noticed your cat lazing in the sunlight with her eyes narrowed down to nothing but vertical slits.  These are both examples of how an iris adjusts to control how much light enters an imaging device – in this case, a person’s (or animal’s) eye.

File:Schematic diagram of the human eye en.svgAs shown in the diagram, the iris (the colored part of your eye) encircles the pupil.  The pigmentation in the iris prevents light from coming through, so light gets in only through the exposed part of the pupil.   The iris is connected to a muscle that controls its size.  When it is dark out, the iris pulls back, exposing more of the pupil.  When it’s bright, the iris relaxes and gets smaller, decreasing how much light comes in. You can think of the iris as a sort of restrictor plate for the eye.

We see when light comes in through the cornea, through the pupil, and then is focused on the retina that lines the rear of the eye.    Cats have a reflective membrane behind the retina that focuses light passing through the retina back into the eye – one reason they see better in the dark.

There are limits to how big or how small the iris can get, which is why we simply can’t see when it’s really bright outside – then we use another type of light restrictor, like sunglesses.  There are also limits to how well we can see in the dark because we can collect only a small part of all the light that is out there.

IrisTelevision cameras as have an iris, to allow them to shoot in a wide variety of light.  Their irises are, as shown, mechanical in nature and are adjusted manually or automatically by the camera based on algorithms that try to optimize the quality of the image.  If you’ve used high quality cameras, you’re familiar with the term ‘f-stop’:  The larger the number, the smaller the iris.  The f-stop (the f is for “factor”) is the ratio between the lens opening and the focal length of the lens.  That’s why f-stops come in such goofy numbers.  The area of the is proportional to the square of its radius.  When you increase the  the aperture by a factor of 1.4, you double the light.  (F-stops are usually 1, 1.4, 2, 2.8, 4, 5.6, 8, 11, 16, 22)

The figure below gives you an idea of how the iris changes the amount of light let into the camera.  The numbers at the bottom are the f-stop values.

Iris_Fstops

The iris on a professional television camera can open way, way up and make it look like it is a lot brighter than it actually is.  It’s simply gathering more light than it would if the iris were at a normal setting.

Twitter was abuzz during the Talladega race with people asking why the drivers were complaining about the lack of light because their television picture looked just fine.  You can’t judge the amount of light from a television picture because the television camera is always optimizing its settings to give you the clearest picture.  You have the same issues with still photographs – how light or dark it looks depends entirely on the setting on the camera or phone.

What I thought was odd were the varying evaluations from the reporters who were actually at the racetrack.  Some were saying there way plenty of light, while others were asking how NASCAR could even think of re-starting  the race given the darkness.  The folks I would really like to hear from are the spotters because if they can’t see their cars, that’s a disaster waiting to happen.

Just for the record, driver reports are similarly unreliable due to a psychological effect that makes you think it’s too dark if you’re leading the race.  If you’re not in P1, the light looks just fine.

Incidentally, installing lighting at Talladega is a tens of millions of dollars project.  Maybe not what a track is able to do given the struggling economy and sagging ticket sales.

NOTE:  As Allen Lee (@wxguy) points out, the CCD (Charge-Coupled Device) – the thing that acts like the retina in your eyeball – is also more sensitive than your retina.   You can read more about CCDs in an article I wrote for Cocktail Party Physics awhile back.

Degrees of Difference: How is Martinsville like Fontana?

 Auto Club Speedway of California, Martinsville, math, Talladega, Texas Motor Speedway, Trigonometry  1 Response »
Oct 252012
 
DLPTXTrackBanking

I love getting questions from readers because I always worry that the geeky stuff I find interesting is only interesting to me.  I love it even more when they not only give me a question, they also supply part of the answer!  This one has to do with the degrees of difference between Martinsville and Fontana.

Michael J. Clark asked a really good question about Martinsville and Fontana:

Why does Fontana (banking in the turns is 14 degrees) seem to have such higher banking than Martinsville (banking in the turns is 12 degrees)?  I would think the 2 degrees more that Fontana has wouldn’t look so dramatically different than Martinsville, but it really does.  I’m guessing it has to do with the fact that Fontana’s turns are about 10 car-widths wide (my estimate) compared to the turns at Martinsville, which seem to be about four car-widths wide.

Great question and another example (like race cars seemingly speeding up when spinning into the grass) of how our perceptions are often subjective.

We always talk about Martinsville being a “flat track”, which is sort of unfair.  It’s flat compared to Talladega and Daytona, but there are still twelve degrees of banking in the turns.  Nothing like a little trigonometry at the racetrack – what does twelve degrees look like?  Let’s start with some definitions so we’re all talking about the same things.

Track width is measured across the track surface and forms the hypotenuse of a right triangle.

Any right triangle can be described by the lengths of any two sides, or the length of one side and one angle.  Remember SOHCAHTOA? You can (finally!!) use your trig to reverse engineer the racetrack.

One degree isn’t really all that large.  A banking angle of one degree means that in order to get a rise of one foot, you need to have a run of about 57 feet.  One degree isn’t very much, as shown in my figure below.

The top picture shows what a banking angle of one degree would look like, with the rise of 1 foot and the run of 57 feet.

The bottom picture is a scale drawing of Martinsville Speedway, which has a track width of about 55 feet (although I think it is a little narrower in the corners).  The banking angle is variously given as 11 degrees or 12 degrees.  I’m using 12 degrees here because that’s what the official NASCAR site says.  Given the hypotenuse (track width) and the banking angle, I can back calculate to show that the rise is about 11.4 feet and the run is about 54 feet.

Now to Michael’s question.  The diagram below shows scale drawings of the banking at Martinsville and California and (just for comparison) Talladega.  I’m using the best numbers I can find on the web.  If someone has more accurate numbers, please let me know.  Kudos, by the way, to Talladega for having one of the best webpages of track data.

Michael has great instincts – the track at Fontana is two-to-three car lengths wider than Martinsville.  This means that the rise is seven feet (one Brad Daugherty) higher than Martinsville at the edge of the track. That increase in rise makes the banking look steeper because you’re looking up a greater distance.

(You always hear that Talladega is five stories tall.  I’m not sure what they’re counting in that calculation because I get 26 feet, which is pretty far short of five stories unless you have very short stories.)

In addition to the greater width, you also have to remember that there’s a huge difference in overall scale.  Martinsville was the second track I visited while writing The Physics of NASCAR – the first was Atlanta.  Martinsville was the track that made me love short tracks.  You get up close to the action and even though they’re not going 200 mph, when you’re that close to them going 100 mph, it seems really fast.  Short tracks are a great challenge to the crew chief (and the driver) because suspension movement is so much more important than aerodynamics.  And, of course, tempers seem to be proportional to the track length of the track:  at Martinsville, they are both really short.

But you have to realize just how much smaller Martinsville is than the California track.  The straights at Martinsville are 800 feet, while the backstretch at Fontana is 2500 feet.  Martinsville is .524 miles, which is 2777 feet.  If you unrolled the Martinsville track, you could just about fit the entire thing on Fontana’s backstretch.  The picture below is my attempt to make a to-scale drawing of the two tracks.  The banking at Fontana looks huge compared to the banking at Martinsville not only because the track is wider at Fontana, but also because the track is simply bigger.  When you look out into the turns, you simply see a lot more asphalt.

Side note:  The featured picture in the post at the top shows me trying to stand up on the 24-degree banking at Texas Motor Speedway, just to give you an idea of how steep 24 degrees actually is.  This was while we were shooting the Science of Speed video series.

So that’s the difference between the tracks at Martinsville and Fontana.  I’m told there is absolutely no comparison between their hot dogs.

Thanks for the question, Michael!  Questions (and suggestions for the Sirius radio “NASCAR Mythbusters” segment) are always welcome.  Click on the ‘contact’ tab above to send me an email.

I’m heading out to Joliet, IL to give my  Science of Speed talk at Joliet Junior College Friday October 26th at 7:00 p.m.  More information on how to get there can be found on my appearances page.  My talk is aimed at the average NASCAR fan and focuses on why it’s a lot harder to drive fast than most of us think.  Most people leave the talk with even more respect for what professional racecar drivers do.  I promise no pop quizzes, so please come on out and meet me!

 

 

NASCAR Concussions II: A Screening Test to Avoid Relying Entirely on the Driver

 Concussion, Dale Earnhardt Jr., NASCAR, Opinion, Talladega  No Responses »
Oct 152012
 

I was lucky enough to speak with Dr. Mark Lovell, an innovator in neurocognitive testing inbetween talks at a conference he was attending.  (Neurocognitive, incidentally, describes those thinking functions that are closely linked to particular areas of the brain.  We’re talking about things like attention, memory, speaking and understanding language, solving problems, and making decisions.)  Dr. Lovell was the founding director of the University of Pittsburgh’s Medical Center’s Sports Medicine Concussion Program and has published over a hundred journal articles and authored or co-authored nine textbooks about sports-related concussions.

Dr. Lovell came to my attention as the developer of the ImPACT (Immediate Post-Concussion Assessment Testing) test, which was one of the tools used to evaluate Dale Earnhardt, Jr.’s concussion.  Given that he is right up the road in Pittsburgh, I thought I’d spend a little time looking into exactly what the test is and whether it’s a viable alternative for NASCAR to use in screening drivers for concussion.

As I mentioned in my last post, physical damage to the brain can be detected by techniques like magnetic resonance imaging; however, even if the physical structure is intact, the brain works primarily on electrical and chemical signals – and we don’t have a way to simply look into a person’s head and verify that all the neurons are sending and receiving signals correctly.

Lovell wanted to make clear that concussions are a rapidly evolving field of research, noting that “90 percent of everything we know about concussion has been learned in the last ten years”.   Ten years ago, no one thought a concussion was a big deal.  Getting your “bell rung” was part of sports like football, hockey and boxing.  We’re learning now — especially from long-term studies of football and hockey players) that even a “mild” concussion in which you don’t lose consciousness can have short- and long-term results.

Here’s what the ImPACT test looks like:  You sit in front of a computer terminal for about twenty minutes and respond to a series of fast-moving activities.   For example, you’re given twelve target words to memorize – you get to see the words for 750 milliseconds each — and you get to see each word twice.  You are then presented with twenty four words, twelve of which are the ones you saw already and twelve of which are similar in some way to the words you saw.  For example, you might be shown the word “ice” as one of the ones to remember and “snow” as a non-target word.    A box comes up and asks “was pencil one of the words displayed?” and you click yes or no.  There are similar exercises using abstract designs, letters, colors, and symbols.  Each exercise tests one (or more) neurocognitive function.

It’s not easy.  In fact, when you’re done, you feel a little like someone has been throwing things at your brain.  The test requires you to multi-task and if you took the time to try to think about how to cheat, it would be obvious that you weren’t giving it your best try.

Lovell emphasizes that a complete evaluation for concussion requires many parts:  The process requires collecting demographic information and a health history survey, including any history of previous concussions.  If you’ve had mental health issues or are taking medications, that has to be factored in as well.  A full description of the injury also is usually included.  Assuming NASCAR adapted this as a standard test, this would all be on file or easily available.

The ImPACT test looks at six distinct elements:

  • Verbal memory (your ability to pay attention, learning and memory relating to verbal things like words, symbols and letters)
  • Visual memory (visual attention and scanning, learning and memory related to visual processes)
  • Visual motor speed (visual processing, learning and memory, and visual-motor response speed)
  • Reaction time (how fast you can respond to something)
  • Impulse control (this is actually a way of measuring the uncertainties inherent in the testing)
  • Symptom score (this measures the presence and severity of 22 symptoms, ranging from sleeping to balance to irritability to fogginess).  This is really the only area of the test in which a driver could try to hide evidence of a concussion.

Dr. Lovell recommends that athletes do a “baseline test” prior to engaging in their sport.  For NASCAR, that would be at the start of every season.  Although the test is still useful without a baseline, having the pre-test allows you to compare a driver’s current state with his previous state and reduces uncertainties that might be caused by things specific to an individual.   Over many years of testing athletes at all levels, the ImPACT people have collected scores for people in different age groups, genders and populations (i.e. athletes vs. us couch potatoes), so even without a baseline, a doctor would have some reference point for how well you ought to score by comparing your score with a typical score of someone similar to you.

One of the challenges of concussion is that it’s a type of injury in which the patient may try to hide the symptoms from the doctor.  Dr. Lovell designed the ImPACT test to help diagnose concussion, even when the subject is reluctant to admit that he or she is exhibiting symptoms.  Trying to purposely score poorly on the test works against the athlete (a bad score makes it more likely you’ll be sat down), so the athlete is motivated to do their best.

Anyone developing a test (whether it be for education or medicine) has to worry about people who try to get around the test’s purpose.  What was most impressive to me in reading the research papers about the ImPACT test is that the developers have come up with ways to tell when people are trying to outwit the test.  Dr. Lovell laughed when he told me that some athletes “sandbag” during their baseline testing, thinking that if they have a lower baseline score, they can get hurt and it won’t be detected.  The test developers have found ways to measure whether someone was trying, for example, to be more accurate by taking the test more slowly, or to complete the test quickly without worrying about being correct.  The test is not only measuring neurocognitive function, it’s measuring whether you’re taking the test to the best of your ability.

Your score on the ImPACT test isn’t a yes/no measure of whether you have a concussion.  Concussion is a subtle enough injury that you still need evaluation by an experienced professional; however, Dr. Lovell believes that the ImPACT test is a good way to screen athletes to determine whether additional medical evaluation is necessary.

Your score on the ImPACT test also gives the diagnosing doctor some help in predicting time to recovery.  The team has developed a cut-off score that broadly predicts whether recovery time will be less than or more than two weeks.  It’s not exact, of course.  Dr. Lovell emphasizes that one of the challenges understanding and treating concussions is that every individual is different.  Most will fall into a “normal” range, but there are always exceptions at both ends of that range.  Repeat testing helps the doctor measure how well the patient is progressing because there can still be neurocognitive impairment well after the obvious symptoms have gone away.

ImPACT is a twenty-minute test administered on a computer.  A system for administering the test could be brought to any race (it’s much more portable than an engine dyno!) and, given that it takes just 20 minutes to complete, seems like a very easy to implement screening tool after any hard impact.  Again, Dr. Lovell would be the first to emphasize that you don’t just take the test and know that you do/don’t have a concussion; however, NASCAR could easily require any driver involved in a crash to do a quick assessment to determine whether they should be seen by a doctor for further followup.

PIt Road Speeding FAQ

 Pocono, Timing and Scoring  14 Responses »
Jun 102012
 

A new record for pit road speeding penalties was set at Pocono this weekend.  Drivers were able to compensate and there weren’t a lot of penalties after the first set.  The question remains:  why all the speeding penalties?

The Facts

Here's the list of pit road speeding (and other) penalti... on Twitpic

There were 10 scoring segments on Pit Road and almost all of the speeding penalties were in the last segment (Pit Road exit).  The photo at left was tweeted by @nateryan after the first round of penalities.  You can view a larger version here.  (Thanks, Nate!)

A few of the drivers shared how much they were told they were over.  Harvick said it was 0.06 mph.  That’s not much.

A Pit Road map showing all the timing and scoring lines was available well before the race.  Ralph Shaheen tweeted the map during the race (Thanks, Ralph!).  NASCAR doesn’t draw their maps to scale (which drives me nuts), so I re-drew the map to scale.  I just realized that I labeled P2 twice.  The first one on the left is P1.  This makes for 10 loops (the start/finish is not part of the pit road timing system).  Note that each of the loops is roughly 207 feet – except the last one, which is only roughly 80 feet.

Questions and Answers

Q. How do they check pit road speeds?
A. Scoring loops embedded in the pit road concrete send an electrical signal when the transponder in the car passes via a process called electromagnetic induction. It’s the same process used to keep track of where the cars are during the race. The transponders are mounted inside the car (I believe they are just a little forward of the drivers’ seat, inside the door.)

The important thing is that they are measuring average speed over the timing segment, not instantaneous speed.  Instantaneous speed is (just as it sounds) your speed at a particular instant.  A radar gun measures instantaneous speed.  If you are on the expressway, you can get a ticket if you go over 65 mph at any instant.  In NASCAR, your average speed is the important thing.  Pit Road speed at Pocono is 55 60 mph.  You can go 50 mph for half the time and 70 mph for half the time and your average speed would be 55 mph.

Q.  Wait – why isn’t pit road speed 55 mph?
A.  Teams get 5 mph over the “official” Pit Road speed.  They all are trying for 60 mph, so let’s not pretend anyone is thinking about going 55 mph.  The Pit Road speed limit is effectively 60 mph.

Q. Do teams know where the loops are?
A. They ought to. It’s not a secret – NASCAR puts out a map showing all the timing lines. Pocono changed the number and location of the timing lines when they repaved Pit Road. It’s hard to believe that a competent crew chief wouldn’t have asked for this information. Given that there was a significant change, it’s also hard to believe that a competent crew chief didn’t discuss this with his driver.  Knauss and a number of other crew chiefs walk Pit Road and inspect all the pit boxes to check on things like cracks, unevenness and little things that might throw off the pit crew.  If they are that attentive to the tiny details, they’d have to have had a major brain lapse not to have asked about the timing lines.

Q. Why doesn’t NASCAR move the scoring lines to keep things interesting?
A. They are embedded in the track. The picture at right shows the loops in Pit Road at Charlotte.  You can’t just pick them up and move them like a garden hose.  If I were repaving a Pit Road, I would play it safe and put in a whole bunch of them, because you can select which ones you want to use for timing and scoring.

Q.  Why didn’t they stop the race and re-calibrate their radar gun after all those penalties?

A1.  Aaragh!  There is no radar gun.  I know people thought it was funny to say because “why don’t they recalibrate their scoring loops” isn’t very humorous.  But don’t saying there is a radar gun because it confuses the people who don’t know.

Would you really want the instantaneous speed to be monitored?  Seems to me that this would officiating a race needlessly complicated.

A2.  I am very skeptical that there was anything technically wrong with the scoring loop system.  It is tested and double tested and the changes of a malfunction are fairly low.  My intuition is that the system was working exactly how it was designed to work.

Q.  Why don’t they just give the drivers speedometers.

A.  Because tachometers are actually more accurate than speedometers.  You can tell your speed to a fraction of a mph with a good tach.

The divisions on a tach are usually 100 rpm. If the driver can read the gauge to 100 rpm, for a typical gear ratio (i.e. let’s say a 1.45:1 second gear and a 4.22 rear end gear), each 100 rpm step on the tach corresponds (for 82.1 inch circumference tires) to 1.37 mph. If you assume that the driver can read the tach to 50 rpm, that’s 0.64 mph.

Besides – when drivers get caught speeding, it is usually because they were trying to cut it too close to the limit, not because they didn’t know how fast they were going.  Remember that you are being judged on average velocity and both the tach and a speedometer measure instantaneous velocity.

I’m not a fan of relying too much on the lighted versions they have now.  I’ll take a dial over a light any day.

Q.  It seems silly to penalize someone for going 0.1 mph over the effective pit road speed limit.

A.  It does; however, where do you stop?  If you tell them you’re giving them another 0.1 mph, then the person who gets caught going 0.16 mph over the effective limit will complain that he was only 0.6 mph over the limit.  NASCAR can’t win on this.  You have to draw a line and it’s going to be arbitrary.  Everyone races under the same conditions, so where the line is placed really doesn’t matter.

Q.  Why don’t they just show us the speeds on the television?

A.  I addressed this before in more detail – just my opinion, of course.  The more data NASCAR hands out, the more they’ve got fans picking apart every aspect of the sport.  I watch races with timing and scoring on my computer, twitter, the radio going over the TV, etc.  I like love data.  At some point, though, you want people watching the race and cheering on their favorites, not picking every millisecond of data apart. It’s a good race for me when I don’t want to look away from the television.

Q.  So why all the speeding penalties?

A.  We don’t know for sure, but here’s my thoughts.

1.  I’m 99.99% confident that the problem was not due to malfunctioning of the system.

2.  TNT advanced the theory that many teams didn’t know the scoring lines had been moved.  Look at the photo and who got nabbed.  The 48 got nabbed twice – once on the pit stop and once while serving the penalty for the first speeding incident.  Of all the crew chiefs who would  would have been on top of the location of the scoring lines, Chad Knaus is #1 on my list.

3.  The last segment – the one that gave so many people trouble – was only 80 feet.  If you’re going 60 mph, you spend 0.94692 seconds in that segment.

a.  At 60 mph, you spend 2.3475 seconds in the long scoring loop segment.  Let’s say you’re going 60.06 mph.  In the short segment, you would spend about one millisecond (one thousandth of a second) less in the segment.  In the long segment, you save 2.3 milliseconds in time.  It would be really interesting to know the accuracy of the system.

b.  A shorter segment is less forgiving.  The drivers are not going at constant speed throughout the loop.  If you push too hard on the throttle for an instant, consider how that affects the average speed if that instant is out of 2.3 seconds or 1 second.  The drivers are constantly trying to figure out how to get maximum speed withing the boundaries of the timing and scoring system.  I’m a little skeptical about the reliance on tach lights – but I’d need to know more about how the lights are set before I could say for sure.  Drivers have many things going on and light may make their lives much easier, but a light is never as accurate as a dial.

NOTE added:  Jimmie Johnson said after the race that NASCAR draws the yellow line at the end of pit road such that the nose of the car is at the yellow line when the transponder is at the scoring loop.  He suggested that the team might want to test that out themselves just to verify the accuracy of the line relative to the transponder.  As I said above, if any team was on the location of the scoring loops, it would be the 48.   I do not believe that not knowing where the lines were was the problem.

NOTE:  Robin Pemberton said on NASCAR Victory Lane that the old segments used to be 274 ft long.

Is Tire Fall-Off the Way to Fix “Broken” Tracks?

 Atlanta Motor Speedway, Brandon Thomas, Bristol Motor Speedway, Bruton Smith, Dover International Speedway, John Probst, Josh Browne, Steve Letarte, Tires  5 Responses »
Apr 062012
 

There’s been an awful lot of talk recently about changing the layout at various track to make racing more exciting.  Bristol is the most-talked-about track, with Bruton Smith planning a $1M revamp of the track to take it back to the way it was before he changed it in 2007.

There are a number of factors that dictate how “exciting” racing is.  For example, the track width and how many “grooves” there are make a big difference in how easy it is to pass cars without “helping” them out of your way with your front bumper.  But last I looked, grip — the source of all speed — is dependent on the interaction of two things:  the tire and the track.  There’s a lot of talk about tracks, but not a lot of talk about tires.

Remember back a few years when tires were a topic of conversation every other week?  Tony Stewart lighting into Goodyear for the tires at Atlanta in 2008?  The Indy tire debacle that same year?  The 2005 Charlotte ‘levigation’ when they “smoothed” the track using a diamond grinder?  Tires aren’t much of a topic these days.  Goodyear’s done an amazing job amidst a slew of re-paving projects from Talladega and Daytona to Bristol and Michigan.

But have they done too good a job?  Some people have suggested that the tires stay in good shape for too long.  It’s possible to go multiple fuel runs without taking tires at many tracks.  If the tires wore faster, might that add an element to the racing that’s missing now by forcing crew chiefs to make tougher decisions about whether to take tires and drivers to take better care of their tires?  Harder tires don’t wear as fast as softer tires – but softer tires are more likely to fail by being worn down rapidly.  It’s a difficult balancing decision and the consequences for Goodyear if they’re not exactly right are significant in terms of how fans perceive the brand.  Take a look at the opinions below and tell me what you think.

Why Does Bump Drafting Seem Much Harder in 2012?

 Aerodynamics, Cooling, Daytona  No Responses »
Feb 252012
 

Although there is a lot of science behind bump drafting, the act of bump drafting is an art.  Even the experienced bump drafters are surprised by the touchiness of the cars this year.

Popping Off: Breaking the Two-Car Draft by Heating up the Engines

 Cooling, Daytona, Engines  2 Responses »
Feb 162011
 

Any closed vessel that is subjected to high temperature will experience increasing pressure.  When that pressure gets high enough, we change from calling it a pressure vessel to a bomb because if(when) it explodes, the vessel itself becomes a collection of high-speed projectiles.  For safety, we don’t heat closed containers if there’s a chance they will reach high enough pressure for them to explode.  A pressure cooker, for example, has a relief valve that at one time was as simple as a rubber stopper tightly fitted into the lid.  The rubber stopper fit in the hole securely enough to handle up to some cutoff pressure, then popped out when that pressure was exceeded.  (This is not an ideal safety mechanism because the flying stopper can injure someone, as can the blast of steam that dislodged the stopper.)

A more practical version is a valve that automatically opens when the pressure exceeds some cut-off value.  The open valve allows excess steam (and sometimes water) to escape.  As soon as the pressure is below the cut off, the valve closes again.  In addition to being safer, it eliminate the time-wasting step of looking for the stopper.

The cooling system on a car is a prime example of a closed system that is heated to high temperature.  Water is pumped through holes in the engine block, where it collects heat.  The now-hot water moves out of the engine and into the radiator, where the heat is transferred from the water to air surrounding the radiator.  The cooled water returns to the engine to pick up more heat.  A Sisyphusian task, indeed.

A radiator is a twist of metal tubing onto which is fastened thousands and thousands of fins that help cool the water that circulates through it.   A typical stock car radiator (like the one at left) might have 20 fins per inch (compared to 10 fins per inch on a typical car radiator).  The more fins per inch, the more surface area available for exchanging heat between the radiator water and the outside air; however, air has to pass through the radiator, so if there are too many fpi, the air flow is decreased and that lessens the cooling.

The water can only carry away so much heat on each trip, so the water temperature gets hotter and hotter as long as the engine keeps producing heat.  The water increases in pressure as the temperature increases.  (See Equation, Clapyron for more on that.)   Water, of course, boils at 212 degrees Fahrenheit, and that would seem to set a limit on how hot you can run an engine; however, there’s a caveat.  Water boils at 212 F only at atmospheric pressure.  As the graph below shows, the higher the pressure, the hotter the water can get before it boils.  Atmospheric pressure is right about 14.7 psi, and that’s where the 212 degree Fahrenheit number applies.  But if you can get the pressure of the system up to about 45-48 psi, the water won’t boil until 275-280 F.  If you can maintain a high pressure in your radiator, you can prevent the water from turning into steam.  Water is much better at carrying away heat than steam is.  Water also flows much better.  Most radiators have a pop-off valve that blows when the pressure gets too high.  A typical radiator cap on a car would be about 15 psi, which actually means 15 psi above atmosphere.  Atmosphere is 14.7 psi, so you’re looking at about 29.7 psi in absolute terms.  This is why your radiator cap has all of those warnings about not removing it while the car is hot:  when the system is vented (opened to atmosphere), the super-hot water will turn into super-hot steam and gush from the opening.

A pop-off valve serves as a ‘weak link’: it has to blow before anything else in the system blows.  Most radiator caps on passenger cars are spring loaded:  When the pressure gets too high, the cap lifts off its seat, opening the system and allowing the hot water to escape into a reservoir.  As soon as the pressure is back down, the radiator cap goes back to being closed.

In a NASCAR car, the pop-off valves open and route the escaping steam and/or water through a tube that passes up near the right-hand side of the car’s windshield.  When you see a car “pushing water”, the maximum pressure has been exceeded and the pop-off valve opened.

For the last couple of years, most of the top NASCAR engine shops have focused on strengthening radiators.  It’s not difficult to get a pop-off valve set to 100 psi.  The problem is that if the pop-off valve isn’t the weak leak in the system, something else breaks.  It’s much more expensive to replace a radiator than a valve – so the size of the pop-off valve is really limited by the strength of the radiator.   A stronger radiator allows a higher pressure to be maintained.  Tim Brewer said that teams were pressurizing their systems to 80 psi (which would be 94.7 psi on my graph were it to extend to the right.)

Two-car drafting produces very high speeds, and that makes NASCAR nervous.  Cutting down the restrictor plate (which they did today) slows down the cars, but NASCAR doesn’t want to change the plate more that 1/64th of an inch or two because the change in plate size significantly affects how air enters the engine.  Teams have been designing engines around a particular plate size, although you would think that by now, they’d know to test not only the announced size, but plates one or two sizes up and down.

The limiting factor on how long two cars can stay in a draft is temperature.  The air intake of the trailing car is blocked when it is drafting, and the water temperature increases.  Two cars could go twenty laps or more before they had to separate.  NASCAR’s plan to limit the two-car draft started with a mandated pop-off valve.  NASCAR requires all teams to use a 33-psi pop-off valve, which corresponds to (33+14.7=47.7 psi) in my graph above.  All the work teams did to manage an 80 psi pressurized system is now out the window.  They also decreased the size of the opening through which air enters the car to cool the engine.  Less air reaching the radiator means less heat transferred from the water and a warmer engine.

Now if someone only could come up with a pop-off valve for drivers…

*****

EXTRA:  Wondering about the different between a tapered spacer and a restrictor plate?  Check out this video, which illustrates very visually how a fluid flows differently through an orifice (the plate) and a nozzle (the spacer).  They’re using both on the Nationwide cars now.  The way the air enters the engine really makes a difference in the combustion dynamics.  Making a smaller spacer would have created too big a perturbation.  The holes in the new space are actually larger, but the plate will help decrease the overall flow.