Results 1 to 8 of 8

Thread: Suspension, as suggested Paul Welsh

  1. #1
    Diamond geezer 500 Club
    Join Date
    Dec 2005
    Location
    East Yorkshire
    Posts
    4,434

    Suspension, as suggested Paul Welsh

    Might put this up to MF if Geoff thinks it not too boring:-

    When larking with the suspension on a new Quik I soon discovered that the criteria in Section S related to energy, and drop - test heights. The real surprise is that it can result in pretty useless (by automotive standards) results. Graphical analysis showed this even more clearly, but that involves maths so lets keep it simple, stupid.

    The ideal relation for the amount of Static Sag, i.e. the amount the suspension settles under standard load is universal, whether car, motorcycle or mountain bike this is around one third of total travel, this leaves one third downwards for hollows and two thirds upwards to cope with bumps. The Quik when measured is nowhere near that standard; in fact I measured two thirds of travel disappear when fully loaded, but we SSDR people are free of all that fortunately. (source of criteria is Mick Broom, who used to make Hesketh motorbikes apparently)

    Another useful titbit is that roughly speaking, if a drop-test height is 260mm and the suspension combined with tyre deflection offers only 58mm of travel, then in theory the 'G' force if dropped from the Section S calculated height is 4.5G, being 260mm total divided by 58mm travel. Since the airframe is designed to cope with 3G negative, even with a 1.5 safety factor, this is of concern. The Quik gets away with at first sight inadequate shocks by airframe flex, but doesn't look ideal...

    My approach is simple and may be too simplistic, but to judge what to allow here, would the following be a valid suggestion?

    1) First work out the drop-test height from the simple wing area and weight sum in Section S; we aren't going to need to drop-test your carefully built aircraft, so don't worry, read on.

    2) Load the aircraft to MAUW and measure from suspension mounting down to the floor, this AUTOMATICALLY gets rid of sag, droop, slack and any other inaccuracies.

    3) Load the suspension attachment locale to 3G (from Section S) and again measure down to the floor, this represent the maximum deflection when landing.

    Item (2) minus item (3) gives total suspension travel and includes tyre deflection; it also neatly combines any play or flex in the system.

    IF the drop-test height divided by total suspension travel is greater than 3G then you have options:-

    1) Beef up the design to the higher figure or better

    2) Increase the suspension travel (or reduce tyre pressure as P&M did)

    Note that the USA ASTM standard asks for nose wheel to be 100 mm higher than Main Wheels, the UK Section S is 'just above Main wheels,' so vague.

    Note also that the UK measure from wheel to floor when calculating drop test height compliance, the reason this is nonsensical is that if the first part of the suspension travel is an over-soft spring, it doesn't count, why?

    What 'S' ignores is simplest physics; the aircraft continues to accelerate UNTIL the suspension and tyre resist with a force greater than the weight imposed upon them. So mandating a drop-test height is meaningless, unless you take into account the point at which the suspension and tyre resist with a reaction force equal or greater than the mass of the machine.

    Once you decide to take the static sag level when fully - loaded, and compare it to the level at the prescribed multiple factor of 3G, you can discover just how much work the suspension is doing to retard your ass from hitting the bump stop.

    And if it isn't enough, you are clearly shown why your airframe or forks are breaking...

    The above checks are really easy to do when considering a 300kilo aircraft, as the numbers are do-able, 450 kilo is more involved though, including sandbags instead of just willing fat pals...;-)

    Cheers

    Kev

    welcoming folk telling me why this is wrong....please?

  2. #2

    Suspension, as suggested Paul Welsh

    I agree that it does need re-visiting Kev.
    But don't hold your breath - a few years ago I submitted an analysis of Betts stitch and Kevlar reinforcing testing criteria, and the huge effect that the tested stitch tension would have on the actual pull into the thread.
    Never had so much as an acknowledgment.
    (However, we do now have a much better guidance, so perhaps it did reach the parts that others didn't - eventually...)
    Dave

  3. #3
    Diamond geezer 500 Club
    Join Date
    Dec 2005
    Location
    East Yorkshire
    Posts
    4,434

    Suspension, as suggested Paul Welsh

    What 'S' is ignoring is that the drop-test height is measured from the bottom of the tyre; if the chosen spring rate is too high or low it takes zero account of this. if they loaded the trike or fuselage to MAUW, measured to a datum local to the suspension mounting, then lifted it to the drop test height before releasing it would make more sense. The wheel can flop or have slack and it would remove inaccuracies.

    The other thing missed entirely is that a mass accelerates under gravity until something reacts with a force greater than the mass; a GT450 or Quik is accelerating well into two thirds of suspension travel, when the shock starts reacting and tyre provides springing to resist. I found that the airframe on the Quik trike series does quite a lot of the energy absorption, luckily the mod of elasticity of alloy is around three times that of steel, so the keel tube makes up for any lacking of the spring rates.

    True deceleration is roughly drop-test height divided by true suspension travel (by 'true' I mean the part remaining after the static sag point); my little trike will have tyre deflection of around 40mm and drop test height is 191mm, so goes through 4.775 G on a drop test, however suspect that the old rule about being able to drop a mouse from any height, and an Elephant from not much at all before killing it applies.

    Cheers

    Kev

  4. #4
    Senior Member 100 Club
    Join Date
    Feb 2015
    Location
    Inside Old Buck ATZ
    Posts
    117

    Suspension, as suggested Paul Welsh

    Kev,

    Your drop test G calculations give accurate results for AVERAGE loads. In practice, the suspension spring / bungee / tyre has a spring rate, which means that there will be considerably more acceleration in the last few millimeters of travel than in the first few. For a linear spring rate, the peak value is double the average value.
    In road vehicles, it is the dampers which smooth out the acceleration by providing a velocity related force. When the wheel first touches the ground, it will start moving at a high velocity relative to the chassis. At high velocity, the damper provides most of the force. As the velocity reduces, the damper force reduces but the spring is becoming more loaded and its force increases.
    A properly damped suspension will reduce the peak forces applied to the airframe to values closer to the calculated average.
    Incidentally, I had understood "negative G" when applied to an aircraft to refer to force generated at the wing, e.g. when flying into an area of strong sink. When landing, the load paths are entirely different and there is no reason to expect the proof load values to be the same.

  5. #5
    Diamond geezer 500 Club
    Join Date
    Dec 2005
    Location
    East Yorkshire
    Posts
    4,434

    Suspension, as suggested Paul Welsh

    No damper fitted Peter; a single-acting damper would not affect the rate of deceleration when dropped anyway. The one flaw in my mod for the Quik suspension was that because it worked almost too well, it had a strong rebound, cured by fitting a check strap as used on the Tanarg.

    That spring mod is on over fifty aircraft last count, and BMAA/CAA required each spring to be load-tested to find out the force required against deflection of the shock unit, then a graph plotted; as the factory spring-rate was too low, by doubling it (effectively stiffening the spring) the deceleration occurs earlier and the deceleration distance measured also doubles. Not intuitive at first glance that a stiffer spring puts less stress on the airframe; my sums showed the drop test and energy results only achieved Section S requirements from deflection in the trike keel, the shock's energy absorption is 32% inadequate when measured alone.

    (see attached graph of stock vs modified)

    Area under the curve = Joules; bigger area = more Joules absorbed. Try driving a soft-sprung family car down a rally track at their speeds and you'll break the chassis, (Citroen DS mastered that trick by long travel)

    Back to Section S

    I had to build a test rig to compress the shock unit and work out all the above, however if you use my simplified method, you no longer have to drop test your carefully built prototype. The aircraft itself becomes the compression rig, and you measure compression at MAUW and compression at minus 3G

    If you design the structure for minus 3G but your suspension is imposing 5G on the airframe when dropped from the approved height, you will have to worry about load paths, the method offers a really simple way to find out what deceleration your suspension and tyre is good for, so you can change tyre pressures and spring rates until the design matches reality, all with 'O' level maths, no more.

    Cheers
    Kev

    Attached files g-june-6th-2011-95mm-red-rubbers.pdf (28 KB)

  6. #6
    Diamond geezer 500 Club
    Join Date
    Aug 2011
    Location
    Kent, UK
    Posts
    1,879

    Suspension, as suggested Paul Welsh

    Kev Armstrong wrote:
    Area under the curve = Joules; bigger area = more Joules absorbed. Try driving a soft-sprung family car down a rally track at their speeds and you'll break the chassis, (Citroen DS mastered that trick by long travel)
    And rising rate spring from the nitrogen filled sphere, as the spring near the end of its travel the spring rate goes up very seriously according to Boyle's Law.

  7. #7
    Diamond geezer 500 Club
    Join Date
    Dec 2005
    Location
    East Yorkshire
    Posts
    4,434

    Suspension, as suggested Paul Welsh

    Spot on Steve

    Stock shock is rising rate but too soft in the early part of travel; it has a slug of P.U. up the centre of the spring which when the spring is compressed expands

    This is clever and mimics a rising - rate, if the spring rate was sufficient (i.e.) double the stock off the shelf unit) the suspension would only sag down one third of travel and work perfectly.

    Rob Mott was really quick on the uptake when I discussed this with him, at 80 kilos static one third sag and 240 Kgf full compression the ideal graph is a straight line, with the internal bump-stop taking care of preventing bottoming

    For an SSDR machine the factory Quik Shock is pretty ideal at 300kilos max.

    I nicked the Morgan sliding pillar idea when making suspension for the Chaser, this was fiddly and too complex as it had to be detachable, in the bad old days of being unable to fly if you changed your machine and it wasn't approved.

    An alloy rod had the spring fitted and slid through a sort of spherical joint, this enabled a lot longer spring for the same overall shock length

    Cheers

    Kev

    Attached files

  8. #8
    Senior Member 500 Club
    Join Date
    Sep 2010
    Location
    Dorset or Mayenne
    Posts
    907

    Suspension, as suggested Paul Welsh

    Kev, I follow this but dropping the elephant on the mouse will kill it from any height.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •