This has been discussed on a previous thread, I forget which.
In a nutshell if you have something to turn, that is able to roll, you can demonstrate to yourself with leggo, meccano or what ever.
First take a wheel and make a hub with a kingpin such that the wheel is as near to the centre of the point of rotation of the king pin. On most two wheel axles this will fall within the tyre area on the ground. Turn the kingpin and the wheel pivots with a slight roll to one way. Return in the opposite direction and it will slightly roll in the opposite direction, this is progression, to allow the wheel to turn some 60 degrees.
In fact most cars have toe in to aid directional stability and caster angle. In other words they will tend to want to go in a straight line, rather than turn. Thus it is that to manoeuvre and protect your tyres you should creep as you manoeuvre, allowing the rolling wheel freedom to take up the angle input progressively, rather than by force, stood still.
None the less the energy required to turn the wheel is not to great and lightens as soon as the wheel is in motion. The movement of the trackrod is short, equal to the degree of turn and does not extend far beyond the kingpin if well designed.
Take the model of the Tri Tech and have some factor of three wheel widths from the kingpin centre of rotation out to where the wheel is. Now the wheel is not pivoting about itself on a spot. It has to physically roll the wheel in the direction of the steering input. It describes a semi circle rather than tiny loop round a point under the tyre. It needs more energy put into the control to do this, and the resistance to movement is greater, not least through leverage that does not exist in the first model.
To acheave full lock the track rod has to travel will past the king pin in a long movement. It, itself, has to describe a slight semi circlular motion to achieve lock to lock. Thus the steering input is proportional, not constant. And suspension movement, and more than likely the different length of track rod to suspension member, create another problem. To achieve up and down movement means a change in the length of steering travel by a proportion, too. So the steering force required to turn is heavier, not constant, and the suspension movement is likely to alter the value of the steering input, in differing amounts, in differing places, as bumps are absorbed. Attempting to smooth out these deflections of direction by steering wheel risks the car loading the suspension the other side of the car and finding the car over corrects. At about 45 mph in a standard Tri Tech the amount of feed back is beyond the ability of the steering control to correct fast and smoothly enough. The car effectively has a steering axle tank slapper and goes out of control. As you approach that point the loadings to correct the steering increase markedly, making accurate input impossible. The imbalanced changing loadings, plus the leverage of the wheels well away from the king pin, continually fight any kind of control. Its a very frightening experience and not unlike a rear blowout on a two/one trike.
The theory that covers much of this is Akerman angle theory. Though on three wheelers it is more difficult to follow. A read through that, and model tests, should provide you with an idea of the possible problems. On the Tri Tech my mod was to increase the track of the car with a spacer mounted on the inner stub axle either side. This made the distance of the wheel to king pin less of the full axle length. I altered the drum and wheels to 10 inch on scooter tyres to bring the wheel centre in board, again shortening the king pin to wheel distance. The car then became drivable at 50 mph and was lighter to drive. The owner seemed happy, but has done few miles.
