Track Building - 

BANKING
How can a banking be designed to avoid cars launching (and crashing into one another)?

Updated December 2003 and August 2004 (changes sidelined)

I've added this section because lots of people read my other articles about track building, and say - hey we want a banked turn in our new track, how do you build it?  What shape should the  banking be and equally important, what gradients should there be on the straights connected to a banked turn?  

Looking at my recent e mails, banked turns are very popular in North America right now.  Thirty years ago many British club tracks had banked turns - some worked well, but many were  plagued with car launching problems on one or more lanes. Perhaps some of the builders at the time understood exactly how to avoid these problems - if they did the knowledge wasn't widespread.  Banked turns have just about died out on British club tracks in the last couple of decades.  These days "flat tracks" sometimes have a few degrees of banking, often associated with gradients around bridges. Usually the angles involved are too small to generate launching problems.  

For the rest of these series of articles I have built (or helped build) tracks using the techniques I've described - so I know they work!  I haven't built a track with high banked turns myself, but Jim Schneider has.  Thanks to Jim both for the photos taken during construction of his track, and for taking the time to check that what I'd worked out in theory does seem to work in practice.

Properly designed bankings shouldn't suffer these launching problems. There are inherent features of banking however clever the design - for example the higher car speeds (higher speeds mean more kinetic energy which increases the risk of damage if it does hit something) and cars being forced down onto the track compressing the rear tyres ( higher g forces are inevitable,  this increases the risk of cars touching down and damaging  the track so a higher ground clearance more rigidly enforced may be advisable).

Material limitations

Slot tracks, including the banked turns, are made of sheet material.  How does sheet material like MDF bend into a curve?  It will naturally bend it into a single plane curve, but bending it into a two plane curve is much more difficult.  There are limits to how tight a bend it will take, and thinner sheet bends more easily than thick sheet.  Perhaps I'd better illustrate the difference between single and two plane curves.  On the left of the picture are a cone and a cylinder which are single plane curves.  On the right are a sphere and a toroid which are two plane curves.  If you want a further illustration of the difference, take a sheet of paper and try bending it into the shapes below. 

Producing a banked bend

A banked bend is really just part of a cone, so the flat sheet can be bent to form a gentle banked turn without too much trouble, steeper banking need rather more force, but the basic shape is still the same.  One way of producing a banked bend is to cut the turn and the straights either end of it from flat board and then pull the straights together.  The diagram below shows how this works.  Note that the initial angle between the straights is increased as the straights are pulled together, and that the radius of the bend is simultaneously reduced.

This technique isn't particularly new,  some 1960's British club tracks  had bankings built in this way.  For example Tony Condon and Dick Smith remember helping build the North London track banking at the old Church Farm headquarters. This banking was built from chipboard, and the straights were pulled together with a rope.  Apparently all that was needed was a lot of tension on the rope and don't worry too much about the groaning noises as it bent!   Once the bending was complete, supports were then added to fix the track surface in it's final position.  As far as  I can remember (I don't suppose anybody can offer a photo - please) it was somewhat steeper than the banking on Jim's track, so it's interesting that "special techniques" to help it bend like the radial cuts or taking it in stages leaving the stresses to relax were not needed. 

  (Click here for more of the story from Dick Smith updated August 2004) 

On the subject of "special techniques" some people have talked about using steam to soften the board and allow it to bend.  Curiously, when asked for more detail, it always turns out this is a story passed on through several people - so I don't know if this is a valuable technique that should be recommended or something that was tried and didn't work (or perhaps somewhere between the two).   Is there anybody out there with practical experience of getting all steamed up could tell us how it really works?

Since originally publishing this article, a builder has suggested that the board should be bent too far, so that  it springs back to the right position rather than still trying to move when it's complete.  I think a little extra pull should work, but how much? One way I can see of predicting how much extra is by experience of building several tracks - any contributions welcome!  The bending stresses in the track surface material will eventually relax, so given time it'll tend to stay somewhere near its new shape. Tracks are not that rigid, so adjustment of the supporting legs can change the banking angles a bit. Commercial builders bond fairly substantial (150 - 180 mm high by 12 mm thick) barriers to the sides of the track, and this seems adequate to hold things in place. I guess they must have some sort of support in place while they do this. I think club tracks often had a wooden framework underneath for support. 

What Angle?

The banking angle that will be produced can be calculated from the initial angle between the straights.  The "What Banking Angle" graph below shows what banking angle is produced when various initial angles are pulled to 180° (that is till the straights are parallel). The following graphs are theoretical and calculated assuming the banking is a section of a mathematically perfect cone.

If you want to calculate it for yourself for any angle, click here.

In practice it's questionable how accurately the  MDF sheet can be persuaded to follow the "perfect" shape.  In fact I'm reasonably optimistic these graphs are a good guide, Jim's track seems to come near to the theoretical values, and I've done small scale tests which come out close to the theoretical values over a range of angles.  I'd be very  interested to hear from anybody else who has tried building a turn this way. 

Also there is a reduction in  bend radius as shown in the diagram (the same circumference of MDF has got to fit round more degrees of bend so the radius reduces). This reduction in radius is shown in the graph below. (for example with a 25 degree banking the graph shows a reduction of 0.9, so if you stared with a flat turn 1m inside radius and 2 m outside radius, the banked turn would have 1 x 0.9 = 0.9m inside radius and 2 x 0.9 = 1.8 m outside radius) 

If you want to calculate it for yourself for any angle, click here.

The straights

This leaves  straights going in and out of the banking which are tipped sideways at whatever the angle of the banking is (see Jim's photo above.  These straights need to come horizontal to fit with the rest of the track. (Unless you are planning an oval with the corners and straight banked!)  The problem that so many banked tracks have suffered from is cars launching.  Cars launching problems often originate from convex gradient transitions, as illustrated below. It is all to easy to produce this sort of gradient transition on at least some of the lanes of the straights either side of a banking.  For clarity, I've  exaggerated the abruptness of the "humps", changes in gradient too gentle to  see in an illustration like this can be enough to launch a car - for example if the hump had a radius of 10m (33ft) a car traveling at over 10m/sec (22 mph) would  launch (and would be very susceptible to bumps at much lower speeds).

A solution this problem is to tip the axis of the banking from the vertical, as shown below.  This means the banking is a bit steeper on the right hand side (half way round the curve) than where it joins the straights.  Here again the angles in the illustration are exaggerated for clarity.  For example a 4 lane track where the straights come flat after 3 m (10 ft), if the track is banked at 20 degrees (angle Y) where it joins the straights, the axis of the cone - angle X - would only need to be tipped at 2 1/2 degrees so the steepest part of the  banking  would  need to be X+Y = 22.5 degrees 

This is a theoretical way of overcoming the launching problem in bankings- I haven't tried building a track like this myself... but I know a man who has.  Look closely at Jim Schneider 's photo of his track (below) and you can see all this geometry on a real track.  In fact Jim overcame the "hump" problem by increasing the banking angle like this from a practical direction before I'd come up with a theoretical explanation.   Obviously the "Concave transition" should be as gentle a curve as possible - if it all happens in a sheet of rather than at the joint between two pieces, the MDF will form a gentle curve naturally.  

Cutting the Slots

Since originally publishing this article, I've had some e mails about cutting the slot.  It wasn't obvious if it should be done before or after bending the flat sheet to form the banking.  It can be seen in the above photos that Jim Schneider bent the track first.  However, the piece from Dick Smith makes it clear that at  North London the slots were cut first, and the surface bent afterwards.

I've had an e mail from another builder saying he tried cutting the slots in flat board, and found he had trouble with the slots narrowing when he bent it into a banking.  This suggests it's a better idea to bend the surface into the banked profile first and cut the slot afterwards. 

It's important to cut the slot accurately, and  more complicated / less rigid jigs are obviously not a good idea. Some techniques (such as a radius arm) work best on flat sheet.  If the slot is cut from the edge then cutting after bending should work fine. An extra complication of cutting on conical surface is that the router will cut shallow (the outer edges of the router bed will touch the track, but there will be clearance under the centre).  This is not a problem for slots which would normally be cut a bit deep anyway, but for braid recesses it may be necessary to adjust the cutting depth in the banked turn.

I know there are other banked tracks being built- so if you trust my theory enough to give it a try, please let me know how well it works out.  This web page could certainly do with some practical feedback!

If anybody out there has found other ways of arranging banking without accidents - I'd be interested to hear all about it!

Chris Frost 
  slot_racing@yahoo.co.uk