Content

View
View Content

Account

Username
Password
Forgotten my password

Pictures

Ready to go
Ready to go

Greenpower Car Frontal Crash Protection - Nose Cone

Rotary Racer, 2015-02-21, Terry Barnaby

Updated for 20g 2015-02-23

These are my thoughts on this subject. I am not an expert in this area, and corrections/feedback is welcome.

The Greenpower challenge is quite a safe activity thanks to the rules and the managing of overall cars speeds. However it does require the designers and developers of the cars and those overseeing their development to make sure a few key safety areas are not overlooked. One of these aspects is roll stability and the roll-bar. This page investigates another important aspect frontal crash protection.

A Greenpower car can be traveling at speeds in the 25 - 40 MPH region in F24 and higher in the F24+ challenge. Race tracks are designed to be reasonably safe with much faster race cars, but there is still the danger of a car running into a solid object, such as a concrete block/wall or another Greenpower car perhaps as a side on impact. With the younger un-trained drivers in F24, this chance is, although low, a possibility and has happened although at relatively low speeds. So the cars need to be designed to try and protect the drivers in these cases.

The main issue is not with the speed but how quickly the speed is reduced, the de-acceleration level. A human body does not like hard de-acceleration or acceleration. The maximum de-acceleration level is open for debate but from various research values not exceeding a peak of 40g, or average of 20g for an adult are typical values.

In the past most Greenpower cars and certainly most of the faster ones had chassis and body work that was relatively yielding. They were made of wood, thin walled tubular metal, thin composite construction etc. most of which would crumple in a crash situation reducing the de-acceleration of the driver and absorbing some of the collisions energy. More recently there has been a push from Greenpower to strengthen the area around the driver to create a more rigid driver protection cell. Although there are some good aspects to this, especially for side impact protection there is a big issue when considering frontal collisions. With a more rigid driver cell there is much more for the front foam nose cone to do in order to reduce de-acceleration to safe levels and absorb the frontal impact energy.

This page gives some thoughts to this and tries to calculate, in an engineering way the size and type of nose cone to handle this.

Maximum De-Acceleration

Colonel John Stapp was the pioneer in research into maximum g-forces a human can survive, much of the research is from his work or based on it.
Formula Student has a rule: "The team must submit test data to show that their Impact Attenuator Assembly, when mounted on the front of a vehicle with a total mass of 300 kg (661lbs.) and run into a solid, non-yielding impact barrier with a velocity of impact of 7.0 meters/second (23.0 ft/sec), would give an average deceleration of the vehicle not to exceed 20 g’s, with a peak deceleration less than or equal to 40 g’s. Total energy absorbed must meet or exceed 7350 Joules."
 
Some random bits of Internet available information:

Calculation of Minimum Stopping Distance

We have to assume some things and define some basic parameters in order to calculate the "impact/crumple zone" needed.

  • Assume the cars structure and seat belts are rigid. (In practice there would be a degree of deformation in these)
  • Assume the car has a frontal compressible area of length L. This would be its frontal impact absorbing structure, the nose cone with a rigid driver structural cell.
  • Now assume a reasonable crash scenario. A car is going at 27 km/H (16.8 MPH) into a solid object. Cars will be going much faster than that, especially F24+, but assume a fair degree of braking. Formula Student bases its calculations on a speed of 25 km/H which is slightly lower, but we have young in-experienced children driving and not so good brakes so I think the degree or braking/avoidance is reduced. This value is open for debate I would like to see the system handle a faster speed than this...
  • Assume your impact absorbing foam/system can compress relatively linearly over  75% of its length.

The crash speed assumption is the most difficult and contentious one ...

Given this:

  1. Speed in meters/second (m/s) is: (27 * 1000 / 3600) = 7.5 m/s.
  2. Average de-acceleration should be no more than: 20g so: (20 x acceleration due to gravity)  = (20 * 9.8) = 196 m/s^2.
  3. So the minimum stopping time is: (7.5 / 196) = 38.2e-3s = 38.2ms
  4. In order to stop from a speed of 7.5 m/s in 38.2ms, assuming a constant de-acceleration force, you will need to stop over a minimum distance of: (0.5 * 38.2e-3 * 7.5) = 0.143m. The 0.5x is due to the car slowing down over the period and assumes constant force.
  5. Now your impact absorbing system has to be longer than this as its compression will become un-linear when getting close to fully compressed so should be at least: (0.143 / 0.75) = 0.190m. This assumes the Impact Absorber produces a constant force over its compression distance and is rectangular and hits something flat..

If instead we assumes a crash speed of 38 km/H (28.75 MPH), the minimum stopping distance would be: 0.379 m.

If instead we assumes a crash speed of 56 km/H (35 MPH), the minimum stopping distance would be: 0.823 m.

For those learning Python there is a small snippet of Python code to do the calculations: crashdistance.py

Based on these figures and leaving a margin for errors etc there should be more than 200mm stopping distance with a constant force rectangular nose cone. To handle higher than 27km/H either a longer nose cone is need or after the nose has compressed then the cars frontal structure should start to collapse to a degree as well to keep the de-acceleration forces low enough. This is of course open to debate...

Impact Absorber

The calculation of minimum stopping distance assumes that the impact absorber (probably a foam block) can absorb the cars kinetic energy by compressing by about 75% of its length. If it is too soft it will "bottom out" and the de-acceleration forces will be much higher if this happens. Also if it is too hard then it will compress less than 75% and the de-acceleration forces will be higher. So we need a suitable material with appropriate compressive strength for the particular shape we have used.

The shape is an important factor. If, for example, it is a 3D cone like shape (as on the front of a squareish cross-sectional  area cars body), there will be very little force required to compress initially (pointy bit) and its full effectiveness will only start to happen once it has compressed by a fair amount.

Different materials offer different levels of resistance to compression and this varies un-linearly with the degree of compression. This is quite complicated but is normally simplified by providing a compressive strength figure which defines an overall force per unit area for some set levels of compression. This is the amount of force needed to compress a unit area of the material by a certain faction of its length and is normally given in Pascals (newton per square metre).

Typically Greenpower nose cones have been made from a black type of foam sold by Greenpower. We don't know its compressive strength so we measured some samples by applying weights to the top and measured how much it compressed by. (Compressive Strength Pa = (weight * 9.8 / area) ) We obtained the following figures (not sure on accuracy ...):

Area Length Weight Distance Distance Ratio Compressive strength
900 mm^2 25 mm 14 Kg collapsed collapsed  
1600 mm^2 25 mm 14 Kg 7.45 mm 0.31 86 kPa
3696 mm^2 25 mm 14 Kg 0.87 mm 0.04  

We think this may be Ethafoam semi rigid packaging foam: http://www.efoam.co.uk/ethafoam-packaging-foam.php which is stated to have a compressive strength of 81 kPa @50%.

We also tested our Polystyrene foam (Dow Floormate 300A rated 300kPa @10% )  that we use for RR's sides:

Area Length Weight Distance Distance Ratio Compressive strength
600 mm^2 25mm 14 Kg 3mm 0.12 228 kPa
1200 mm^2 25 mm 14 Kg 1mm 0.04  

The Formula Student challenge specifies a particular impact absorber to be used made of Dow Impaxx® 700 and with a particular size/shape.

Material Dow Impaxx® 700
Datasheet https://www.rollbarpadding.com/FS/CO/84/0/IMPAXX700.pdf
Compression strength at 50% 23 degreesC 835 kPa

Dow Impaxx 500

Material Dow Impaxx® 500
Datasheet https://www.rollbarpadding.com/FS/CO/111/0/IMPAXX500TDS.pdf
Compression strength at 50% 23 degreesC 612 kPa

Dow Impaxx 300:

Material Dow Impaxx® 300
Datasheet http://www.dow.com/scripts/litorder.asp?filepath=automotive/pdfs/noreg/299-51012.pdf&pdf=true
Compression strength at 50% 23 degreesC 434 kPa

Now some calculations. Assume:

  1. Cars all up weight is:120 Kg
  2. De-acceleration required is: 196 m/s^2
  3. Assume an impact absorber with a rectangular shape (not aerodynamically shaped).

Average force needed to stop car at 196 m/s de-acceleration is (from F = ma): (120 * 196) = 23.5 kN

Area of Dow Impaxx® 500 foam needed: (23.5 / 612) = 0.038 m^2 (about 195mm x 195mm)

Area of Dow Impaxx® 300 foam needed: (23.5 / 434) = 0.054 m^2 (about 232mm x 232mm)

Area of Greenpower foam needed: (23.5 / 86) = 0.27 m^2 (about 522mm x 522mm).

Some Internet resources:

Impact Absorber Shape

The above calculations assumed a rectangular impact absorber shape. Often on a Greenpower car the nose is rounded or pointed for aerodynamic reasons (often not so important ones ...). If the nose cone shape is more pointy then a greater length of foam is needed to absorb the energy at the de-acceleration rate required in the distance required as initially the de-acceleration force would be lower than the average force needed

How to calculate this ....

  • Use reduction of foams Volume reduction due to shape as a scaling factor ?
  • Using this a 2D Hemisphere shape would thus require an extra length factor of:: (4/PI) = 1.27. So a 200mm requirement would be: 254mm long.
  • This needs some more thinking ...

Conclusions

This just covers the basics and uses basic Mathematics and ignores some of the finer details, but I hope is about right.

In order to reduce de-acceleration to less than 20g from 27 km/H on Rotary Racer, a typical Greenpower car, the following would probably be an minimum (assuming chassis compresses/ starts to collapse at higher than 27 km/H speeds):

Car shape 600mm wide x 400mm height so rectangular overall cross-section looking from front
Material Dow Impaxx® 300
Shape Roughly hemisperical shape rounded front in 2D (wideish car shape)
Width 400mm
Height 140mm
Length 250mm

This assumes hitting something flat on rather than say a narrow pole and at a speed of less than 27km/H with a max g Force of 20g.

There are questions on what crash speed and maximum de-acceleration to aim for. 27 km/H (16.8 MPH) sounds a bit low for the nature of the drivers in GP. There are also no thoughts as to how the seat-belts and frame would affect things and especially how the relatively heavy (for the child's size) helmet would effect things.

A 100mm foam nose cone, assuming has spot on the correct material for compression to 75% with the car in question, would only be ok for up to a 10MPH collision.

A sideways impact into another car is perhaps the most dangerous and most likely possibility. In this case the sideways impacted car would be knocked sideways with roughly half of the acceleration force (assuming it is of the same weight and hit at the center of mass). The acceleration forces on each driver would be half in this case and could thus handle double the speed as the other car would start to move at 1/2 the speed) (Is this right ?). One issue would be if the side impact can be resisted by the car that is hit.

To lower the de-acceleration and  to cope with a higher speed impact a greater length of foam of smaller area or lower compressive strength is needed.

As pupils are now learning the Python computer programming language, here is a very simple piece of Python code to perform the calculations given here. crashdistance.py