On magnets and weights in slot cars
by Andrew Rowland
This study is directed specifically at ‘classic’ F1 models marketed by Scalextric (SX) in the last few years, although it may be that some or all of the results would equally (and possibly more easily) be applied to Saloon type cars too.
My belief is that the magnets inserted by the factory in these cars cause undesirable results in several ways:
So the first thing I did was simply remove the magnets, but sadly that didn’t have the desired affect. These cars (I speak now exclusively about the SX ones) were clearly NEVER tried or tested or indeed intended to be used without the magnets. Which is a great pity since it really doesn’t need much work or study to either reduce dramatically the strength of the magnet so that the cars still have most of the above desirable characteristics or, even more simply, to make them heavier so they perform much like the original cars.
Take for example the CARTRIX cars, their magnets are subtle and make a much reduced degree of difference, allowing ‘tuning’, not total change to performance.
More importantly still CARTRIX magnets are easily removed and replaced without causing any potential damage to the car, something that it is very, very difficult to achieve with the SX models which are so complex and fragile that after I had taken out the magnet (which necessitated taking out the motor and therefore in turn the rear axle assembly) doing this only once had caused stress fractures to appear in the axle mounting clips meaning I am now loathe to ever take the model apart again since I risk breaking it.
Whilst I appreciate these models are ‘scale’ rather than full ‘racing’ and that they are sold primarily for their fine detail and not their strength, it still seems to me incorrect to sell something as a ‘functioning’ electrical toy with screws holding it together (and not simply glued) that cannot actually be taken apart without terminal damage being done to it. These new cars seem to have been made with a very brittle plastic and the forms are far thinner than in the past and these factors don’t seem to have any logical reason that I can see.
Anyway this study is of another type, so to continue……
More specifically I was interested in whether the performance change brought about by weights was in some way ‘predictable or tuneable’ in order that different models could have ‘similar’ performance in order that they could compete against one another.
As a study I have chosen 7 cars. These are split into 3 groups for the purposes of the test:
(Three of the four have been chipped but the Maserati 250FI have found impossible to locate the chip inside without damaging the driver, something which I am currently loathe to do. I would be interested to hear from anyone who had indeed managed this, to understand how it might be done. In fact I also have the tinplate version and this I have managed to chip so I may end up using that to complete the set of four.)
(Sticklers for authenticity will tell me that actually the Mercedes 196 never raced the Vanwall 56/57 but since I’m English, I live in Italy and have German friends I like the international nature of these four cars on the track and it is only 1 year…...)
This seems to be mainly due to the tyres and the extreme lightness of the model.
Notes / Assumptions:
So pitting the CARTRIX against the SCALECTRIX in Group 1. is already hard since the SCALEXTRIX cars obviously have scaled tyres while the CARTRIX R’n’D dept. obviously decided that the performance needed to be enhanced and so gave their cars tyres that are slightly wider (and softer).
More than just width or diameter however the fundamental thing is the compound, the material or the softness. I’m not an expert on this so I cannot say whether the two companies tyres are indeed a different material, however what I do know is that despite the two SX cars being of exactly the same wheelbase, track and almost the same weight and also given they are both Goodwood revival cars of the same limited edition I would expect them to have the same tyre compounds.
Actually I was quite a long way into my study before I realised that the tyres on the Maserati were much harder than those on the Vanwall and that is what had always made them so different to drive. This really pissed me off frankly! It meant however hard I worked on my results for this study I will never be able to race those cars together since their performance is so different. In fact the Maserati, without the magnet is almost as undriveable as the Lotus 49!
I actually ended up using only the soft tyres (originally from the Vanwall) and changing them between the two cars for the study, but is it only me that thinks it is totally ridiculous that two cars have two different tyres?
Build quality / fragility of the models
Track type / length affecting the relative
performance of each car
All I can say on this is my track is what one might call ‘average’ with some R2 curves and one ‘long’ straight, so I suggest my findings are a good base for others to follow in most cases.
Variation between models / real cars
Having said that I nearly always lose but again as someone once it’s the competition that is important not the winning.
Car Weights and speeds
This gives some idea of how the much lighter newer cars despite all the modern guides, motors etc. don’t actually necessarily make for better performance.
As a control I also raced a few original SX cars to try to understand what sort of lap time might be ‘fun’ and reasonable as a target. Obviously this ‘desired speed’ is in some way subjective but actually there are several things dictating that speed:
I originally decided my desired lap time would be 3.9seconds, although later this was changed to 4s as I realised that 3.9s was not achievable by some of the cars. Below the tabulated times are therefore quoted at 4.0s while some of the early results regarding the Vanwall were based on achieving the 3.9s lap.
Using a simple spreadsheet I worked out the ratios of weights (dry to magnetic) and the lap times (again dry to magnetic) and assuming a directly proportional relationship between weight and time this allowed me to calculate the weight of the car at any lap time I chose.
Of course (as was immediately pointed out to Graham when he aired my idea on the Forum) this is very simplistic, given that, it doesn’t allow for inertia. In fact, as I will explain below there are other things it doesn’t consider too, which are also fundamental to a car’s lap time.
Initial Data Collection:
* predicted by reference to other models since to take it apart and insert the magnet would almost certainly have caused terminal damage to the axle mountings.
** in practice I assumed an average of 196gr for these two models since one had a broken magnet and was probably showing a spurious reading. This might have been a wrong assumption since the Ferrari may have been intended to have a more powerful magnetism to overcome its reduced rear end track width.
*** these times were taken with the infamous ‘hard’ tyres and since I was not going to take the risk of dismantling the motor again to put the magnet back I had to ‘extrapolate’ the results for calculation use.
**** it should be noted that this lap time is BELOW the desired result of 4s. This has been ignored for the purposes of the study so that this car is assumed to perform at the desired 4s and left as is. If it is seen that this car is too consistently quick I may add weight to the nose to reduce rear end traction.
Adjusted Data Used for calculation:
The consequent final weight predictions for a 4s lap time, together with the amount therefore required to add are thus:
* one should recall this 4s represents a better performance in real terms than Group 1. since the track had by now been lengthened by around 4% and a ‘chicane’ curve added meaning that these cars would need to be better for the same lap time than Group 1.
I simply took some roofing lead of 1.5mm thickness and cut it up with cutters to the desired shape. Where necessary to achieve more precise shaping I used a scalpel but any craft knife would work adequately.
For the Lotus 49 I also used other tools (see below).
Group 1. test Results
Since the car is chipped it wasn’t easy to fit weight in, however I managed to try different solutions all the way up to 99gr (an addition of 38gr!).
Further I thought that putting the weight behind the rear axle was going to make the rear want to rotate out since it would create a larger ‘moment’ in respect of the pivot point of the car (the guide).
Thirdly it is generally acknowledged that the weight should be as low as possible to help prevent the car ‘toppling’, through raising the centre of gravity.
Several solutions were tried and the results were:
Obviously I have listed them here in order of weight in order that a pattern emerges although in reality the tests were done rather more randomly than that until the best result was achieved.
Conclusions to Vanwall test:
Obviously point e) could at least partially be overcome through the use of a tuneable controller of the ‘Slot.it’ type whereby the braking could be set to maximum however this is beyond the scope of my simple study.
Overall it became immediately clear that my initial calculations were nothing more than a starting point, so from the point of view of the initial supposition that weights could be used in a predictable way to ‘simulate’ the magnet it was clear that the experiment was not a success, however from the point of view of driveability the car really had been transformed.
In a subjective sense:
It was seen that this performance window really ‘popped’ for me at 83gr, although one should consider that this is always a 1950’s car of very narrow track with 3mm wide tyres so it should be borne in mind that this car will never handle like a modern F1 with wide slicks.
Drivers used to another kind of car might still take a while to adapt to this kind of driving.
One final point to consider is that the original Vanwall C 55 (round pin) will do a very similar lap time with its old RX motor, pinion and drive cog but with new tyres, which in my view shows how fabulous those old cars were and how all the advances in technology haven’t actually got us forward very far!!! (Here it should however also be considered that the original car was actually about 1/30th scale not 1/32nd making the track, length, tyre width etc. rather larger and therefore the car is consequently easier to drive.)
For the record this was 17gr on the flanks, 4gr behind the driver and 8gr in the rear, making a total weight of 84gr against the 85gr prediction.
I was immediately disappointed that the lap time was still some 0.5s above the expected 3.9 / 4.0s.
It was only at that point that I stepped back and questioned why two cars with the same wheelbase, track and almost the same weight had such totally different performance levels. It was then that I noticed that the tyres were quite simply of different colours and on inspection therefore of different compounds.
This upset me quite a lot and after I swapped the tyres over and tried the Maserati with the soft tyres it immediately went 0.4s quicker per lap, lowering the time to about 4.1s.
Conclusions to Maserati Test:
The only other thing to note is that the long tail and height of the available space on the Maserati means that extreme car needs to be taken not to put too much weight there to avoid the car being too ‘tail happy’.
Group 2. test Results
Thinking back this wasn’t necessarily clever since the Ferrari was quite a lot slower and more difficult to control probably due to its much reduced rear track.
Anyway, I managed to pack in 20 of the predicted 25gr necessary and this brought the lap times down from 4.4s to 4.1s, a big improvement although not sufficient to compete with the original ‘contemporary’ SX models, but at least representing a better lap time than the 1950’s cars.
The slightly modified and extended track was deemed to be about 0.3s longer than the old one such that the original 4.0s lap time was now about 4.3s. Consequently these cars were lapping about 0.3 / 0.4s quicker than the 1950’s ones in Group 1. a minimum when you consider their low centre of gravity and wide tyres (I assume aerodynamics not to have much affect at this tiny scale).
How this was achieved can be seen in the photo and this represents about the maximum achievable without removing the motor (again) to put lead in where the magnet was although I would recommend anyone taking out the magnet to do this at the same time so as not to have to remove the motor twice which is a major hassle and risks destruction of the extremely beautiful but delicate detail.
This wasn’t actually that difficult and necessitated 25gr of the 34gr prediction to get down to 4.0s.
My only final comment on these is that they still handle like dogs! They sound dreadful (all rattling along, probably due to the front wheels being located separately) and slide all over the place.
Again, I suggest that the tyres are simply not up to the job without the magnet pulling them down hard onto the track.
Finally I noticed last time I took off the top body part that the motor ‘airbox’ has a large empty space in it which sits pretty close to the rear axle (at least as close as much of the weight which I have added until now), so think it might be a good idea to fill that up with lead.
It wouldn’t be impossible to match this increase in the Ferrari as there are still untapped cavities inside there, so it might be possible to get those times down and the handling up.
If anyone wants to try and let me know I’d be happy to hear about it………
Group 3. test results:
Actually I should also say that I had (and still have) the intention to weight the Ferrari 156 Sharknose and the Cooper T53 but that their dimensions and the difficulty therefore of adding weights has put me off a little, however the challenge of the Lotus 49T was simply too much, especially after the gauntlet was laid down by the initial response to Graham’s posting on the Forum.
Following the experiments with the other 4 cars above I had become somewhat adept at ‘knowing’ what was required so less experimentation was going to be required here. That said this model offers ‘special’ problems due to the fact that the monocoque really has been made just big enough to accommodate the motor and driver (just like Colin Chapman designed the real thing!).
Initially I played with taping weights to the underside of the body since there is about 1.5mm of space all the way along (except where the magnet housing and the gearbox poke down). This was actually a good starting point as it enabled me to test various weight quantities and distributions rapidly, something impossible with the other models since they have to be reassembled every time to test them.
Again several solutions were tried between 6 and 17 gr. It should also be remembered that the initial weight addition prediction required was 11gr.
Again this did not provide good driveability or significantly better lap times (I didn’t even record them at this stage).
Now we really had a problem since there really isn’t anywhere in that car to put any more weight. I was already unhappy with the one underneath although I guessed you wouldn’t really ever see it with the car in motion.
(Now I know that this is officially the Lotus 49T (‘T’ for Tasman), not the European GP car but as I understand it apart from the fact that the engine had a capacity of 2.5l rather than 3l they were essentially the same car.)
That decided I cut a crude shape of wing and tapped it to the top of the gearbox to see how that would perform.
Extraordinarily it was a vast improvement. Still pretty slippery but absolutely what was needed.
The two weights brought the car up to 67gr some 18 above the original and 7gr above the prediction but this seemed to work well. With this set up I achieved a 4.4s lap pretty quickly and this was very ‘repeatable’ making for a good predictable ‘race able’ car. The quickest lap achieved was 4.3s.
Given the original lap time was 4.6s, plus the fact that (as we saw above) a probable lap difference from old to new track configurations represents about 0.3s this would equate to an old lap time of 4.0s, making this car at least comparable to the Group 1. cars and a total of 0.6s a lap quicker than ‘dry’.
The odd thing about this car is that unlike the Vanwall / Maserati the weighting made quite a small difference to the lap times despite the huge increase in driveability. This seems odd since one would think that driveability would quite quickly be translatable into control and therefore time saved. I suppose that given sufficient practice this might be the case, once one had adapted sufficiently one’s driving style but under my limited lap regime this didn’t occur.
My personal conclusion is that this car suffers from the repeated tyre compound problem. Once weighted the car was more driveable but still had far more power than traction making wheel spin inevitable under acceleration. Plus the motor sensitivity is just too high so that the slightest squeeze of the trigger creates too much power output (much like the real car at the beginning I believe). Again this must be tuneable with a better controller.
So I looked again at various Lotus 49 solutions and happened upon a photo of Jochen Rindt at Monaco in 1970. In this form the rear spoiler was quite large, low and situated directly above the rear axle (as was logical in those early days of spoiler design).
So……. I soldered one up out of the lead sheet and with a couple of spacers to keep it level and off the gearbox and a couple of ‘prongs’ to locate it between the radius rods and suspension springs I had a weight of 15gr sitting exactly above the rear axle! (Plus its removable for adjustment / modification or showing the original car off.)
So now my Lotus 49 goes pretty well (despite the tyres) and weighs 64gr and I have to say it looks pretty good too (judge for yourselves) with the wing which obviously I need to paint and refinish to make it look better.
Of course one can easily criticise the crude soldering and the ‘attachment’ is not particularly robust but I’m going to rework the fixing when I have time in the future and try to further ‘finish’ it once I’ve got the undercoat on.
Another option I looked at but didn’t execute would be the Monaco 1968 version which had a sort of flat inclined plate fixed over the engine. I don’t think, unless this was more than 1.5mm thick, this would alone provide all the weight necessary, meaning the under gearbox weight would also need to be employed. This would however be easier to model being essentially ‘flat’, if one felt the wing design was too complex and / or not sufficiently ‘beautiful’.
It also occurs to me that there is a business here for anyone who can fabricate a simple mould to cast these things in lead. Experiment might show that a few grams less would offer a more beautiful casting (especially if the side plates could be thinner), and this would be a great compromise.
Now, I understand that most English readers don’t have access (yet) to the Gold Leaf Lotus and obviously the earlier ‘BRG’ version would look unrealistic with wings but if I managed to buy a kit from Spain from Italy, you could have done too! Anyway, it looks like they’re coming eventually so you’ll just have to wait a little longer!
One final detail is that I now have to add the front side wings to ‘complete’ the car and I’m toying with the idea of doing this in lead rather than plastic just to add a few more grams of weight. I’ll do some tests first and let you know. The sad thing about this is that finally I’ll have to modify a car, since until now I’ve managed to do all of the above without once ‘damaging’ the original model.
Well, that’s all I have time for, I’ve got quite hot so I’m going to take my anorak off!
I hope someone out there enjoys what I’ve written and I’d be so happy to hear from anyone who’s done anything similar. If anyone’s got ideas how to further enhance performance on these cars I’d also be interested to hear and if anyone’s managed to chip the Maserati, Cooper T53 or Ferrari 156 I’d be REALLY interested to hear about it.
AR - November 2009
|Lo-Fi Version||Time is now: 10th December 2016 - 02:59|