How to define steering deadband zone?

[jani12]

Consider a hydraulic steering system with small electric motor atop. The maximum output torque of this motor is about 2 N.m. This output torque is input to worm gear. The output torque of the worm gear is about 20 N.m.

This steering subsystem is mounted in a prototype truck.

The requirement is that this steering subsystem must not provide steering assist while in steering deadband zone. When the handwheel is turned in deadband zone, there shall not be impact on the road wheels.

How to define steering deadband zone? For example should it be -5 deg to 5 deg or what should it be?

 

[crutschow]

How to define steering deadband zone? For example should it be -5 deg to 5 deg or what should it be?

How would we know that?
It's your design so you need to specify how many degrees of deadband you want/expect the system to have.

 

[jani12]

What was the rationale for choosing deadband range on some of the projects?

 

[Tony Stewart]

Deadband is defined by +/- % max steering torque which translates into +/-deg on steering wheel where driving wheels change with some low threshold. The rationale is to spare the pump at highway speeds for long term small adjustments thus reduce the duty factor of pump operation, yet not cause noticeable tension bumps to driver.

 

[rjenkinsgb]

with small electric motor atop. The maximum output torque of this motor is about 2 N.m. This output torque is input to worm gear. The output torque of the worm gear is about 20 N.m.

That's not practical.

A normal worm gear cannot be "back driven" so a failure would lock up the steering.

A failure of any power boost or assist system should only cause reversion to unassisted manual steering, not cause total failure; that's just not permissible.

 

[Tony Stewart]

Deadband is defined by +/- % max steering torque which translates into +/-deg on steering wheel where driving wheels change with some low threshold. The rationale is to spare the pump at highway speeds for long term small adjustments thus reduce the duty factor of pump operation, yet not cause noticeable tension bumps to driver.

Measure PSI in one direction to mid-position. Subtract PSI in reverse until motion is detected = deadband PSI
% deadband = deadband PSI / total range of PSI

 

[Tony Stewart]

That's not practical.

A normal worm gear cannot be "back driven" so a failure would lock up the steering.

A failure of any power boost or assist system should only cause reversion to unassisted manual steering, not cause total failure; that's just not permissible.

Cars with worm gear steering return towards straight after a corner and no load on steering wheel. This is reverse torque.

 

[jani12]

small electric motor atop. The maximum output torque of this motor is about 2 N.m. This output torque is input to worm gear. The output torque of the worm gear is about 20 N.m.

Let me re-state. Worm gear is used for torque multiplication. There is 1:10 gear ratio.
For example, if motor output torque is 2 N.m., after torque multiplication, torque at steering column or input shaft is 20 N.m.

 

[Tony Stewart]

That's not practical.

A normal worm gear cannot be "back driven" so a failure would lock up the steering.

A failure of any power boost or assist system should only cause reversion to unassisted manual steering, not cause total failure; that's just not permissible.

I don't understand your reverse torque concern.

 

[rjenkinsgb]

Cars with worm gear steering return towards straight after a corner and no load on steering wheel. This is reverse torque.

Vehicles with the main steering gear using a worm system generally use recirculating ball or similar low friction units, which can be back-driven easily with near zero loss from either side.

That is very different to trying to add an electric motor and 10:1 worm to drive the steering column itself.
(Note I did explicitly state a "conventional" worm drive - which cannot be back driven).

Also note the main steering on this is specified to be using a hydraulic power steering system.

Consider a hydraulic steering system with small electric motor atop.

A failure of the motor control would lock the steering solid, if such a thing was added.

An override for eg. lane keeping on such a system would usually be either additional hydraulic control, or a direct-drive torque motor added to the column, which has no significant influence or drag when unpowered.

 

[shortbus=]

But, non rack and pinion steering is not really a worm gear in the common sense of the word. The use some thing called "worm and sector". The sector takes the place of the common worm gear "worm wheel".

A steering worm and sector is different too in that the worm ratio changes from the center section to both ends.

https://www.researchgate.net/figure/Worm-and-sector-steering-gear-system-http-wwwtpubcom-basae-213htm_fig1_325923660

 

[JimB]

Look here:

Marles Weller Steering Gear (P12) - The MG Automobile Company

All parts comprising this steering assembly are extremely rare now. There is a good opportunity that those items numbered can be supplied. RECONDITIONING SERVICE: A full reconditioning service is available for this type of steering box. Please ring or write for full details of time taken, price...

mg-parts-spares.co.uk

My first car, a 1935 MG PA had a Marles-Weller steering box with a spiral "worm wheel" (item 4) which engaged with a peg (item 11).

The steering would self centre with the castor action of the suspension, and manually moving a front wheel would turn the steering column/wheel no problem.

It took about 1.75 turns lock to lock on the steering wheel,
power assistance? not required, beautifully light steering on 4.50 x 19" tyres.


JimB
(wishing he still had the car)

 

[Tony Stewart]

Vehicles with the main steering gear using a worm system generally use recirculating ball or similar low friction units, which can be back-driven easily with near zero loss from either side.

That is very different to trying to add an electric motor and 10:1 worm to drive the steering column itself.
(Note I did explicitly state a "conventional" worm drive - which cannot be back driven).

Thanks for the heads up. I didn't know some worms were called self-locking.

snip
"All theoretical analyses of self-locking worm gears deal with static conditions.

In such an analysis, the load on the worm gear can’t drive the worm if the coefficient of friction between worm gear and worm is larger than the tangent of the worm’s lead angle. In other words, the friction angle must be larger than the lead angle to prevent backdriving."

 

[rjenkinsgb]

I didn't know some worms were called self-locking.

That's what I'd consider to be a conventional worm drive - metal on metal, no special feature to reduce friction between the worm and wheel.

Jim's example has bearings set in the follower "pin" for friction reduction; it's not a design I'd seen before, but has the feature to allow low-loss back drive.

This is a style I'm more familiar with (my older brother is a Jaguar specialist..) - Recirculating ball, effectively a ballscrew & nut system:

 

[JimB]

Jim's example has bearings set in the follower "pin" for friction reduction

Just a quick note, the bearings in the pin are not spherical balls, but hemispheres.
The flat side of the hemisphere contacts the worm, and the round side fits into a round recess in the pin.

JimB

 

[shortbus=]

Another thought on "dead band". Don't know much about the new electric steering, but the old hydraulic power steering the so called dead band was set by the steering valve built into the steering box. The valve only supplied power after a certain amount of torque was applied to it from the steering wheel/shaft.

 

[Beau Schwabe]

shortbus= ... reverse thinking of that, if you have some overlap in the hydraulic system, then they are BOTH supplying power that cancels out (or simply recirculates) within the deadband.

 

[shortbus=]

shortbus= ... reverse thinking of that, if you have some overlap in the hydraulic system, then they are BOTH supplying power that cancels out (or simply recirculates) within the deadband.

The only problem I can see with it working like that is that it wouldn't allow the steering to self center after a turn.

 

[rjenkinsgb]

The only problem I can see with it working like that is that it wouldn't allow the steering to self center after a turn.

That is actually the normal operation with hydraulic power steering.

The pressure, return and both cylinder ports are all open to each other (port overlap) when there is no torque on the column, so nothing counters the centering force from the suspension geometry or blocks the power cylinder from being moved.

Torque on the column from the steering wheel progressively closes off one or another set of ports, to direct pump flow through the actuating cylinder in the appropriate direction.


There is no true "dead band" where movement has absolutely zero effect, it's just that small movement add little bias to the flow so very small force imbalance at the power cylinder.

That is progressively increased as more torque is applied and the port areas allowing free flow get smaller, until with sufficient steering torque all pump flow has to go in one direction through the power cylinder.


Example of one type of column valve; different types have different numbers of port sets, but are generally similar

Another example in this page:

AGCO Automotive Repair Service - Baton Rouge, LA - Detailed Auto Topics - How Power Steering Works

AGCO provides overall lowest costs of vehicle ownership. We do this by providing extremely high quality automotive service, following the theories and practices of Dr. W. Edwards Deming., how hydraulic power steering works

www.agcoauto.com

 

[shortbus=]

That is actually the normal operation with hydraulic power steering.

I should have added that no pump pressure is on both sides of the cylinder at rest. it is mostly residual pressure. As shown in your second link.