Joined
·
3,296 Posts
SUV Suspension Design (2002)
by Bob Sheaves
Neither Allpar, its affiliates, nor Robert W. (Bob) Sheaves accept any responsibility for the use of any information contained in this article. This information is supplied for educational purposes only. Readers are advised to seek competent assistance before attempting to use this information in any way.
Bob Sheaves wrote:
Utility vehicles can be grouped into several subcategories, which must be accurately defined to prevent confusion and misinterpretations. These categories are defined by weight class (as a side note, they have nothing to do with state and federal licensing or weight laws):
This page only talks about vehicles in Class 1. In a PBS program about the rollover history of the 1992-2000 Ford Explorer and 1983-92 Ford Bronco II, comparisons were made and conclusions drawn based upon incompletely presented information. This report will attempt to address the missing information.
Ford's "Twin Traction Beam/Twin I-Beam" independent front suspension:
In the late 1950s, Ford determined the need for a smoother ride and handling closer to passenger cars in its Class 1/2 pickup line, to pass Chevrolet and GMC as the largest business purpose class vehicles in use. Several designs were looked at, from trunion type (as used on later AMC front suspensions), to more softly sprung Hotchkiss designs (see below for description of this design), to fully independent A-arm designs of various configurations. Marketing determined that the appearance of a fully independent suspension would lead the customers to believe that the IFS was not as "strong or durable" as the Hotchkiss style solid front axle, with its parallel high rate leaf springs.
Analysis of the frame showed that the frame would flex incorrectly (unlike today's vehicles, the frame was, at that time, seen as part of the total suspension system, and these components were designed to flex, intentionally) and the hard points of the A-arm design would lead to ground contact patch instability. The ground contact patch is defined as the area contacted by the tire at ground level at the tire static loaded radius. This instability of the GCP (Ground Contact Patch) in the A-arm design would lead to unpredictable and potentially unsafe handling in emergency maneuvers.
To uncouple the left and right sides of the vehicle (to provide a better ride quality with no loss of control), Ford engineers came up with the "Twin I Beam" (which is a rear wheel drive ONLY design) concept, where the idea of a lower A-arm is expanded into a pair of crossing beams and angled control arms, forming an A-arm, mathematically, of around 36 inch span between frame mounting points, and an apex height (at the kingpins) of approximately 24 inches. This single "control arm" would have the steering knuckle with a .75 inch diameter kingpin. The knuckles were tied to the steering box by using an inverted "Y" linkage, where the long arm attached to the steering box pitman arm on the left side of the vehicle (inside the frame rail) and the other end was connected to the right hand steering knuckle.
The left steering knuckle was attached to the drag link (previously connected to the right knuckle) by a tie rod attached to the left knuckle running to the right side of the vehicle and attached to the drag link approximately 12 inches from the right hand knuckle attachment of the drag link. This design approximates the swing arc of the control arm, to reduce the toe change inherent when the suspension cycles from the jounce (full "up" position) to rebound (full down or "droop" position) limits.
Springing was handled by either leaf or coil springs depending on the weight classification of the vehicle. Rubber bushings provided jounce control (when mounted to the underside of the frame, allowing a positive stop for the control arm to hit and thereby limiting jounce travel) and the extended shock absorber provided the rebound limit control over the downward travel.
Roll control in corners was generally handled (on coil sprung vehicles) by a separate torsion reaction "anti-sway" bar, while leaf sprung vehicles relied on the torsion effect of the leaf itself to provide this control.
Hotchkiss style (CJ/YJ) non-independent front suspension:
The Hotchkiss design suspension has been virtually unchanged since the early 1900s when semi-elliptical springs were introduced to replace the fully elliptical springs used on the Model T and other vehicles.
This design consists of a pair of parallel, semi-elliptical leaf springs rigidly attached to the axle by (usually) u-bolts, minimizing pitch laterally and the resultant misalignment of the axle to the direction of travel.
Additionally, the leaf springs are generally mounted such that there is 1/3 of the overall chord length mounted towards the fixed point mounting, and the remaining 2/3 of the chord length mounted towards the pivoting shackle, allowing the chord length to "grow" as the spring flattens out under jounce loads.
Roll stiffness (resistance to body "roll" in a corner) in this design is roughly 2 to 3 times the stiffness of a coil sprung design, depending on factors not discussed in this report, due to the interest in brevity.
Steering is affected greatly by the manufacturers choice of steering linkage designs. Most common today is the Haltenburger design, due to the use of panhard, or "track" bars to limit lateral torsion in the leaf springs. This is the design utilized in the Jeep YJ.
In the YJ design, the panhard rod is mounted to the left frame rail and attached to the axle tube on the right side, inboard of the knuckle yoke. The Haltenburger linkage follows this same basic layout, with the drag link attaching to the pitman arm on the left, and the other end connects to the right hand steering knuckle. The left knuckle attaching point tie rod end is connected to the drag link approximately 12 inches away from the right hand knuckle.
The important factor in this design is that the angle of inclination between the drag link and the panhard rod MUST be as close as possible to parallel and equidistant at all steering input angles. Mismatch of this angle will cause "head toss" and jerky steering in the vehicle.
Conventional unequal length upper and lower A-shaped control arm independent front suspension:
A former Jeep/Truck engineer (not Mr. Sheaves) wrote, "The 2005 Grand Cherokee went away from the link-coil front suspension, as most people know, partly for smoother on-road manners and because of pressure from auto writers.
A big reason for the switch, which is less well-known, was [for the factory] to be able to stuff the entire powertrain and suspension up into the body in a fraction of a minute.
The XJ, ZJ, and WJ were laboriously assembled with the links, springs and axles coming up from below; the engine, transmission, and transfer case were swung in on a hook from above, as you would do at home (including the banging and cursing). The WK update was a boon to manufacturing efficiency."
Unequal length control arm suspensions are the most complex of the designs considered, but also, potentially, the most "tunable" thanks to the isolation and adjustability of the various dynamic parameters, such as anti-dive, roll control, anti-squat, travel, etc. This very complexity entails analysis and determination of desired handling characteristics far more demanding, however. All of the manufacturers of Class 1/2/3/4/5 trucks have utilized this basic design over the years, but cost of the components has continued to be a justification stumbling block.
The term "unequal length" comes from the fact that, by design, the pivot length of the upper control arm (UCA) is shorter than the lower control arm (LCA). This is measured between the pivot axis and the ball joint, forming the triangle or "A" shape. Usually, the LCA is the longer of the 2 arms and travels through the lower measurement chord arc height. The UCA, being shorter, travels through a shorter chord arc height. (This means the chord arc height is measured horizontally.) The upper control arm moves the top of the tire a greater distance laterally in the vehicle than the lower control arm moves the GCP. This means, in a dynamic condition, the the tire wobbles at the top a lot more than the ground contact patch. This instability is what must be controlled by the geometry design to prevent a vehicle roll-over condition and directional instability.
Safety and SAFETY
As an suspension design engineer, I have to serve 3 masters:
Of these, the end consumer is the hardest to satisfy, as most are not capable of telling me, in engineering terms, what they look for in a vehicle. My interpretation of these wants, and the vocational needs of a vehicle, may be on target (as on the Dodge "Baby Ram" Dakota pickups, when introduced in 1994) or off by a mile ("King of the Hill" Jeep mule - a 4x4 with independent front and rear suspensions).
Generally, the Program Manager will have a list of design requirements that are key to the program, and are easy to meet. "Engineering good design practice" will trip the unwary.
Look back at the Ranger in mid-flight. This looks to be unsafe, right? Nope - from an engineering standpoint this vehicle is actually pretty good, because it does exactly what the builder wants, and is repeatable from jump to jump. This is not to say I agree, but simply that it meets the builder's requirements. Structurally, it is also safe, being built from 4130 chrome-molybdenum steel tubular framing, instead of cheap stamped HSLA "C" channel frame members. It also incorporates a full 12 point rollover protection cage and 5 point occupant seat belt, shoulder harness, and anti-submarine belt for each occupant-tested at a far higher standard than the MVSS your passenger car is tested to.
The suspension is designed such that a 50 mph evasive maneuver will cause the truck body to roll to approximately 7 degrees (by comparison the stock Explorer only is allowed to roll about 5 degrees). Put another way, the Ranger prerunner body will tilt 3 more inches than the Explorer, about the roll axis. Depending on the engineering parameters used, this is either good or bad, depending on what the design intent was initially.
COMMENTARY
For the above reason, I have to laugh when I see people saying that "this or that design is unsafe." There are no absolutes. Naderites need not respond.
The driver of a vehicle is FAR MORE responsible for control of that vehicle that I am as a suspension designer. There are no such things as "an accident". There is ALWAYS a reason for something happening-whether it is politically acceptable or not to admit it.
Did you know that AMC (and Chrysler) did not lose even ONE case in court over the "roll overs"? In EACH case, alcohol, drugs, lack of training all played a factor in the death of these people. At the risk of seeming uncaring (and to quote Dr. Jerry Pournelle in the book "Oath of Fealty"), "think of it as evolution in action."
END COMMENTARY
Money also raises its head during the design process. Costs often get too far out of hand and compromises must be made to accommodate the amount of money the consumer is willing to spend, the amount of money in the budget for that component subsystem, and the amount of money to be spent on design time. There has never been a car designed that is a theoretical "perfect" car. The "best" or "perfect" car for one person is a "piece of mud scraped off my shoe" to someone else. It is the design engineer's job to provide the most value for the target audience.
When regulators get involved, without the necessary background to truly understand the issues (meaning they simply pander to the "home crowd" to get re-elected), the vehicle design engineer is placed at a distinct disadvantage-simply put, the more flashy you are, the more the press will cover you. This AIN'T a design engineer...LOL! The downside is that second guessing, ignorance, and lack of general concern allows myths (such as the CJ roll over "problem," see above) to perpetuate.
CONCLUSIONS
To sum up all this rambling into a few sentences:
I have been involved with automotive design and engineering since 1975, when I first started working for IH/Scout Racing. My personal specialty is vehicle dynamics, handling, and suspension design. The previous generation (1994-2000) BR (full size Dodge Ram) pickup 4x4's are all my "kids" as I was responsible for 4x4 suspension design on that vehicle when I was an employee of Chrysler Corp, PreProgram Engineering Group (Advanced engineering group located at JTE in Detroit).
I recently "retired" from the industry when I returned home to the US after working in Japan for Isuzu and other OEs for over 2 years. Currently, I am doing some consulting on road racing car suspension modifications and axle/driveline design on the side, but my "job" is being a father to my new son, Michael, born last December.
Notes on steering mechanisms in cars
Recirculating ballcan be just as viable and precise as a rack-and-pinion, due to the physical limits of attaching the inboard tie rod ends to the gear itself. There are only just so many positions available, without creating a new gear housing, while a recirculating ballcan be placed virtually anywhere needed and connected with a relay rod to the centerlink. Toe control is far easier and more precise when you can link to the proper positions. I'll take a recirc gear anytime when I can get the geometry I need to create those high lateral forces without understeer or oversteer. A custom gear from TRW or ZF will cost over $2.5M to design, develop, tool, and manufacture. ... the positive reason to go to a rack and pinon are sprung weight redustion, less complexity, less tolerance stackup inherant in the design. The debits are geometric compromise when the rack to tie rod end attachment do not meet the needs of the suspension, cost, loading tolerance.
Home •
Engines •
Reviews •
Chrysler 1904-2018 •
Upcoming •
Trucks •
Cars
Spread the word via <!--Tweet or--> Facebook!
We make no guarantees regarding validity or accuracy of information, predictions, or advice - .
Copyright © VerticalScope Inc. All rights reserved. Dodge, Jeep, Chrysler, Ram, and Mopar are trademarks of Fiat Chrysler Automobiles.
by Bob Sheaves
Neither Allpar, its affiliates, nor Robert W. (Bob) Sheaves accept any responsibility for the use of any information contained in this article. This information is supplied for educational purposes only. Readers are advised to seek competent assistance before attempting to use this information in any way.
Bob Sheaves wrote:
Utility vehicles can be grouped into several subcategories, which must be accurately defined to prevent confusion and misinterpretations. These categories are defined by weight class (as a side note, they have nothing to do with state and federal licensing or weight laws):
- Class 1 - 0 to 6500 lb. GVW
- Class 2 - 6500 to 10,000 lb. GVW
- Class 3 - 10,000 to 36,000 lb. GVW
- Class 4 - 36,000 to 80,000 lb. GVW
- Class 5 - over 85,000 lb. GVW
This page only talks about vehicles in Class 1. In a PBS program about the rollover history of the 1992-2000 Ford Explorer and 1983-92 Ford Bronco II, comparisons were made and conclusions drawn based upon incompletely presented information. This report will attempt to address the missing information.
Ford's "Twin Traction Beam/Twin I-Beam" independent front suspension:
In the late 1950s, Ford determined the need for a smoother ride and handling closer to passenger cars in its Class 1/2 pickup line, to pass Chevrolet and GMC as the largest business purpose class vehicles in use. Several designs were looked at, from trunion type (as used on later AMC front suspensions), to more softly sprung Hotchkiss designs (see below for description of this design), to fully independent A-arm designs of various configurations. Marketing determined that the appearance of a fully independent suspension would lead the customers to believe that the IFS was not as "strong or durable" as the Hotchkiss style solid front axle, with its parallel high rate leaf springs.
Analysis of the frame showed that the frame would flex incorrectly (unlike today's vehicles, the frame was, at that time, seen as part of the total suspension system, and these components were designed to flex, intentionally) and the hard points of the A-arm design would lead to ground contact patch instability. The ground contact patch is defined as the area contacted by the tire at ground level at the tire static loaded radius. This instability of the GCP (Ground Contact Patch) in the A-arm design would lead to unpredictable and potentially unsafe handling in emergency maneuvers.
To uncouple the left and right sides of the vehicle (to provide a better ride quality with no loss of control), Ford engineers came up with the "Twin I Beam" (which is a rear wheel drive ONLY design) concept, where the idea of a lower A-arm is expanded into a pair of crossing beams and angled control arms, forming an A-arm, mathematically, of around 36 inch span between frame mounting points, and an apex height (at the kingpins) of approximately 24 inches. This single "control arm" would have the steering knuckle with a .75 inch diameter kingpin. The knuckles were tied to the steering box by using an inverted "Y" linkage, where the long arm attached to the steering box pitman arm on the left side of the vehicle (inside the frame rail) and the other end was connected to the right hand steering knuckle.
The left steering knuckle was attached to the drag link (previously connected to the right knuckle) by a tie rod attached to the left knuckle running to the right side of the vehicle and attached to the drag link approximately 12 inches from the right hand knuckle attachment of the drag link. This design approximates the swing arc of the control arm, to reduce the toe change inherent when the suspension cycles from the jounce (full "up" position) to rebound (full down or "droop" position) limits.
Springing was handled by either leaf or coil springs depending on the weight classification of the vehicle. Rubber bushings provided jounce control (when mounted to the underside of the frame, allowing a positive stop for the control arm to hit and thereby limiting jounce travel) and the extended shock absorber provided the rebound limit control over the downward travel.
Roll control in corners was generally handled (on coil sprung vehicles) by a separate torsion reaction "anti-sway" bar, while leaf sprung vehicles relied on the torsion effect of the leaf itself to provide this control.
Hotchkiss style (CJ/YJ) non-independent front suspension:
The Hotchkiss design suspension has been virtually unchanged since the early 1900s when semi-elliptical springs were introduced to replace the fully elliptical springs used on the Model T and other vehicles.
This design consists of a pair of parallel, semi-elliptical leaf springs rigidly attached to the axle by (usually) u-bolts, minimizing pitch laterally and the resultant misalignment of the axle to the direction of travel.
Additionally, the leaf springs are generally mounted such that there is 1/3 of the overall chord length mounted towards the fixed point mounting, and the remaining 2/3 of the chord length mounted towards the pivoting shackle, allowing the chord length to "grow" as the spring flattens out under jounce loads.
Roll stiffness (resistance to body "roll" in a corner) in this design is roughly 2 to 3 times the stiffness of a coil sprung design, depending on factors not discussed in this report, due to the interest in brevity.
Steering is affected greatly by the manufacturers choice of steering linkage designs. Most common today is the Haltenburger design, due to the use of panhard, or "track" bars to limit lateral torsion in the leaf springs. This is the design utilized in the Jeep YJ.
In the YJ design, the panhard rod is mounted to the left frame rail and attached to the axle tube on the right side, inboard of the knuckle yoke. The Haltenburger linkage follows this same basic layout, with the drag link attaching to the pitman arm on the left, and the other end connects to the right hand steering knuckle. The left knuckle attaching point tie rod end is connected to the drag link approximately 12 inches away from the right hand knuckle.
The important factor in this design is that the angle of inclination between the drag link and the panhard rod MUST be as close as possible to parallel and equidistant at all steering input angles. Mismatch of this angle will cause "head toss" and jerky steering in the vehicle.
Conventional unequal length upper and lower A-shaped control arm independent front suspension:
A former Jeep/Truck engineer (not Mr. Sheaves) wrote, "The 2005 Grand Cherokee went away from the link-coil front suspension, as most people know, partly for smoother on-road manners and because of pressure from auto writers.
A big reason for the switch, which is less well-known, was [for the factory] to be able to stuff the entire powertrain and suspension up into the body in a fraction of a minute.
The XJ, ZJ, and WJ were laboriously assembled with the links, springs and axles coming up from below; the engine, transmission, and transfer case were swung in on a hook from above, as you would do at home (including the banging and cursing). The WK update was a boon to manufacturing efficiency."
Unequal length control arm suspensions are the most complex of the designs considered, but also, potentially, the most "tunable" thanks to the isolation and adjustability of the various dynamic parameters, such as anti-dive, roll control, anti-squat, travel, etc. This very complexity entails analysis and determination of desired handling characteristics far more demanding, however. All of the manufacturers of Class 1/2/3/4/5 trucks have utilized this basic design over the years, but cost of the components has continued to be a justification stumbling block.
The term "unequal length" comes from the fact that, by design, the pivot length of the upper control arm (UCA) is shorter than the lower control arm (LCA). This is measured between the pivot axis and the ball joint, forming the triangle or "A" shape. Usually, the LCA is the longer of the 2 arms and travels through the lower measurement chord arc height. The UCA, being shorter, travels through a shorter chord arc height. (This means the chord arc height is measured horizontally.) The upper control arm moves the top of the tire a greater distance laterally in the vehicle than the lower control arm moves the GCP. This means, in a dynamic condition, the the tire wobbles at the top a lot more than the ground contact patch. This instability is what must be controlled by the geometry design to prevent a vehicle roll-over condition and directional instability.
Safety and SAFETY
As an suspension design engineer, I have to serve 3 masters:
- End consumer
- Vehicle Program Manager
- "Engineering good design practice"
Of these, the end consumer is the hardest to satisfy, as most are not capable of telling me, in engineering terms, what they look for in a vehicle. My interpretation of these wants, and the vocational needs of a vehicle, may be on target (as on the Dodge "Baby Ram" Dakota pickups, when introduced in 1994) or off by a mile ("King of the Hill" Jeep mule - a 4x4 with independent front and rear suspensions).
Generally, the Program Manager will have a list of design requirements that are key to the program, and are easy to meet. "Engineering good design practice" will trip the unwary.
Look back at the Ranger in mid-flight. This looks to be unsafe, right? Nope - from an engineering standpoint this vehicle is actually pretty good, because it does exactly what the builder wants, and is repeatable from jump to jump. This is not to say I agree, but simply that it meets the builder's requirements. Structurally, it is also safe, being built from 4130 chrome-molybdenum steel tubular framing, instead of cheap stamped HSLA "C" channel frame members. It also incorporates a full 12 point rollover protection cage and 5 point occupant seat belt, shoulder harness, and anti-submarine belt for each occupant-tested at a far higher standard than the MVSS your passenger car is tested to.
The suspension is designed such that a 50 mph evasive maneuver will cause the truck body to roll to approximately 7 degrees (by comparison the stock Explorer only is allowed to roll about 5 degrees). Put another way, the Ranger prerunner body will tilt 3 more inches than the Explorer, about the roll axis. Depending on the engineering parameters used, this is either good or bad, depending on what the design intent was initially.
COMMENTARY
For the above reason, I have to laugh when I see people saying that "this or that design is unsafe." There are no absolutes. Naderites need not respond.
The driver of a vehicle is FAR MORE responsible for control of that vehicle that I am as a suspension designer. There are no such things as "an accident". There is ALWAYS a reason for something happening-whether it is politically acceptable or not to admit it.
Did you know that AMC (and Chrysler) did not lose even ONE case in court over the "roll overs"? In EACH case, alcohol, drugs, lack of training all played a factor in the death of these people. At the risk of seeming uncaring (and to quote Dr. Jerry Pournelle in the book "Oath of Fealty"), "think of it as evolution in action."
END COMMENTARY
Money also raises its head during the design process. Costs often get too far out of hand and compromises must be made to accommodate the amount of money the consumer is willing to spend, the amount of money in the budget for that component subsystem, and the amount of money to be spent on design time. There has never been a car designed that is a theoretical "perfect" car. The "best" or "perfect" car for one person is a "piece of mud scraped off my shoe" to someone else. It is the design engineer's job to provide the most value for the target audience.
When regulators get involved, without the necessary background to truly understand the issues (meaning they simply pander to the "home crowd" to get re-elected), the vehicle design engineer is placed at a distinct disadvantage-simply put, the more flashy you are, the more the press will cover you. This AIN'T a design engineer...LOL! The downside is that second guessing, ignorance, and lack of general concern allows myths (such as the CJ roll over "problem," see above) to perpetuate.
CONCLUSIONS
To sum up all this rambling into a few sentences:
- The Ford Explorer, while not my personal favorite design or even second, is a safe vehicle in my opinion.
- The Firestone tires, coming apart as they did, showed massive underinflation. This is undeniable. The only question remains is WHO set that pressure? This ultimately boils down to the consumer following the manufacturers recommendations. Just what those recommendations were, I cannot comment on as I have no first hand knowledge of them.
- The final conclusion is that the vehicle does not enter the safety issue until FAR AFTER the driver. If the driver is not capable of understanding and demonstrating this understanding, of the handling characteristics of his or her chosen vehicle, that driver has NO business behind that steering wheel.
I have been involved with automotive design and engineering since 1975, when I first started working for IH/Scout Racing. My personal specialty is vehicle dynamics, handling, and suspension design. The previous generation (1994-2000) BR (full size Dodge Ram) pickup 4x4's are all my "kids" as I was responsible for 4x4 suspension design on that vehicle when I was an employee of Chrysler Corp, PreProgram Engineering Group (Advanced engineering group located at JTE in Detroit).
I recently "retired" from the industry when I returned home to the US after working in Japan for Isuzu and other OEs for over 2 years. Currently, I am doing some consulting on road racing car suspension modifications and axle/driveline design on the side, but my "job" is being a father to my new son, Michael, born last December.
Notes on steering mechanisms in cars
Recirculating ballcan be just as viable and precise as a rack-and-pinion, due to the physical limits of attaching the inboard tie rod ends to the gear itself. There are only just so many positions available, without creating a new gear housing, while a recirculating ballcan be placed virtually anywhere needed and connected with a relay rod to the centerlink. Toe control is far easier and more precise when you can link to the proper positions. I'll take a recirc gear anytime when I can get the geometry I need to create those high lateral forces without understeer or oversteer. A custom gear from TRW or ZF will cost over $2.5M to design, develop, tool, and manufacture. ... the positive reason to go to a rack and pinon are sprung weight redustion, less complexity, less tolerance stackup inherant in the design. The debits are geometric compromise when the rack to tie rod end attachment do not meet the needs of the suspension, cost, loading tolerance.
Home •
Engines •
Reviews •
Chrysler 1904-2018 •
Upcoming •
Trucks •
Cars
Spread the word via <!--Tweet or--> Facebook!
We make no guarantees regarding validity or accuracy of information, predictions, or advice - .
Copyright © VerticalScope Inc. All rights reserved. Dodge, Jeep, Chrysler, Ram, and Mopar are trademarks of Fiat Chrysler Automobiles.
