1969 MGB GT Build Log

It has been a while since an update but quite a bit has happened. I took some time off to rest as well as visit family, we were staring down wildfire evacuations, work has been very busy, and I had my wisdom teeth removed leaving me out of action for a couple days. In that time though I have kept quite busy with car things as well.

First up was getting a mockup of the engine/transmission location in the engine bay scan in CAD to see what I was working with. I compared a few different transmission scans, a BRZ, 5 speed Miata and 6 speed Miata and there are some interesting observations I posted in a different thread. The main ones are the improved bellhousing and mounting situation on the BRZ transmission, I posted some overlays if anyone is curious. To confirm the selection of the BRZ transmission I needed to confirm I could design a kinematic linkage that would place the shifter where I wanted and once I worked that out, I made sure clearances with the transmission tunnel looked good. I haven’t designed a shifter mechanism, but I was able to show that kinematically it should be possible with minimal non-linearity and no binding. The BRZ transmission has an adapter plate available from kPower to make the mate-up simple.

I found a JDM BRZ transmission locally for a very good price with very few miles and pulled the trigger, at the same time I put in an order for the kPower adapter plate.

At the same time, I started work on the front suspension kinematics. I started by cleaning up the scan data I have of the front s2000 upright and building out the corner assembly model. I took a profile of the inside of the wheel and worked out rotors and caliper selection as well. It seems like the narrow Wilwood Dynapro’s should be workable with a Mini Cooper S front rotor. Finally, I confirmed the piston sizing in my excel tool and I am happy with where I am bias wise, and my ability to make changes if necessary.

I moved the upright kinematic points from CAD into the kinematics software. To make the lack of a rear ARB work well I forced the roll center to the ground in the front which also minimizes jacking and means I will need some amount of bar in the front giving me some tuning of the relative roll stiffness. With that set and some other basic things like keep out zones, a first draft of the front-end kinematics came together. This is only the first draft though and it will be edited once parts are designed and clearances, part availability and stiffness are considered.

At the same time as working though the kinematics I started addressing the steering rack. The stock MGB steering rack is actually a very good option and hits most of the targets. I need a rack with a width pivot to pivot of around 24 inches, a fast ratio of around 2 inches per revolution and a total travel of around 6 inches. Most importantly though, the column needs to be very close to the center of the rack to clear the very narrow frame rails on the car. This turned out to be quite difficult to find and other than the MGB rack almost nothing else would work.

There were two main candidates other than the MGB rack, a BMW Z4 steering rack and a Mustang II rack. The Z4 rack has the correct ratio, is a manual rack, has the correct width and near the correct travel. Additionally, the thread on the end of the rack would let me use off the shelf Honda CRV inner tie rods to connect the rack to the OEM S2000 outer tie rod ends at the correct with. The only problem is the column leaves the rack about 0.75” farther away from the center and collides with the narrow frame rails.

The Mustang II rack is also a good option and meets all the criteria except for the speed. There are simply no racks up near 2 in/rev and therefore a steering quickener box would need to be used which doesn’t make me feel great.

The MGB rack is a good option, but it has one major drawback. The inner tie rods attach to the rack in a slightly unusual manner and adapting a longer inner tie rod, especially one with a metric outer thread would be quite hard. In addition, new options don’t seem very high quality and even the expensive option from an otherwise reputable supplier seems to be yet another version of the same rack that is widely available. This leaves me with the final option of machining the ends of the existing MGB rack and rebuilding what I have. I don’t think this will be too big of an issue, but I don’t personally have a lathe so I will have to find a friend who can help. I tore down the rack I have which required a big fight to remove some of the backlash mechanism but the major components look in rebuildable shape.

To help with final engine placement and understand the steering rack clearance I lifted the engine into the engine bay and compared the position with the CAD. I had some concern over the alternator clearing the frame rail as well as whether the default manifold would work facing rearwards. To my surprise, neither was an issue and I will be able to use the OEM water pump, alternator and intake manifold/throttle body. The intake and throttle body I have are the high flow versions from the JDM Type R K20 which are hard to get.

Once the bellhousing adapter arrived, I bolted the engine to the transmission and there are a few pieces of hardware I will need to buy but otherwise it fit nicely. I had no issues with the slave cylinder or starter motors clearing. I do still need to re-test the engine and transmission in the engine bay for clearance and I think a few areas may now pose a problem but for the most part no modifications will be required to make it fit.

With the engine largely fixed in place I started work on a subframe for the V0 geometry and it is starting to take shape. I don’t have any images of it yet, but it looks a fair amount like a much smaller version of the rear subframe. I am still working out some details and making sure that nothing will penetrate the cockpit in an accident, but the basic geometry is coming together. I’m trying to use a rear attachment point in addition to the normal mounts as the standard setup is compliant, or heavy.

I put together a complete project plan for the front suspension over the next 5 months and if everything goes well the car should be a rolling chassis by the middle of the year sometime. This time I let MS Project handle the scheduling and I am getting more and more proficient at using it to manage a complex project.

Finally, I moved the chassis down to floor level and bolted up the rear subframe to the car. There are some adjustments I would like to make to the rear mounts on the chassis which are only tacked in place now, but largely it fits well. I need to make upper damper mounts and then I can put the car down on it’s own wheels for the first time in over 3 years. Nothing in life goes perfectly though and impatience got the better of me by bending the brake caliper tabs on the RHS front upright while pressing out the wheel bearing. I’ll put in another order for that, and life will carry on. Sorry for the massive update, I’ll try to get back in the habit of smaller, more digestible updates moving forward and past the holiday season.

Outstanding, as always!

Running into a bit of a mental block. I spent some time in the garage this weekend going back and forth between CAD and the real thing with the engine and transmission in the car and jigged into it's final position. Both now agree and I am happy with all the clearances. The bottom of the oil pan is 5 inches above the ground at ride height which was about as aggressive as I wanted to push it, and ensures that the accessories also clear the frame rail. Additionally I received a RHD AE86 quick rack which through some miracle I discovered solves all my problems.

The problem is reacting the LCA forces properly into the frame without either intersecting the engine which I can't go under, interfering with the steering rack which is largely in place or putting a whole ton of tubes or plates in inefficient loading conditions like bending. I don't currently have a tubing bender so small bends to cheat past things aren't really possible. In the configuration below the front LCA pickup point is able to efficiently transfer lateral forces and longitudinal forces and while it technically is able to transmit vertical forces it shouldn't see too many of those. The front also clears all the things it needs too and provides convenient mounting locations for the steering rack and eventual ARB, all fine and dandy.

The rear is another story though, that has me pretty stumped. My early designs used the original MG subframe mounting locations and a tube across the car to react the balanced lateral forces under braking. Additionally I had mounts similar to the front mounts on the rear subframe that grabbed the frame rails under the car. Effectively it was a smaller, but similar version of the rear subframe in the front. My complaint with that design though is the occupants limb safety in an accident. I don't feel like having strong metal bars shoved into my feet and legs in any collision large enough to push the subframe back. Not being able to brace across the car well due to there being an engine in the way and the low location of the pickup points relative to the frame rail make it a bit of a tough problem to address any forces in the lateral and vertical plane.

The screenshot below obviously has all the tubes that support the rear points removed and only the tubes required to manage the front. My current ideas have all involved brackets down to the frame rail that are stiffest laterally and react the vertical forces through bending. Does anyone have any ideas they can sketch out on the image below that constrain the rear mounts, don't pass across the engine and don't point tubes at the occupants feet? For reference, the height of the longitudinal tubes is the same as the frame rails on the bottom of the car.

Could you redesign your lower control arm using  a rear strut rod that mounts to the existing frame rails eliminating the need for your new subframe to extend that far to the rear?

if you have a tube bending service near you it might be worth it to just get one tube cut.

is modifying the sump out of the question?

Not sure if you are worried about NVH but if not you could use the engine or trans as a stressed member, a lot of dirtstock cars out here bolt the engines to the frame through the transmission adapter plate.

I appreciate the ideas! I mulled over this for a few days and had some busy days at work where I could come back to this with fresh eyes. Seems like bending is the way to go, sometimes there just is a right tool for the job. I need to run some studies to see where I am stiffness target wise and to make sure I'm not missing anything big, and then update the kinematic model with the small changes I made to confirm it hasn't introduced any new problems. One final check of the CAD after that and I'll be good to start working through the mounts and then onto control arms.

Since the last update I have confirmed all the kinematic geometry, put together the majority of the subframe in CAD and made some progress on upper and lower control arms. After doing analysis on the front subframe the magnitude of the forces under braking became clear. Images of both the straight braking and combined braking and cornering are below. The loads are simple added together, higher than in reality. The key difference is some loads are canceled out and some become unbalanced.

In the rear of the car the forces are fairly small, a relatively small thrust force when accelerating applied at the center of the wheel and a small braking force applied at the tire/ground contact point. At the front the majority of the braking force the vehicle can produce is applied at the contact patch which is a much larger force and also applies a large torque to the assembly. This results in very large forces in the rear member of the lower control arm. Additionally, to allow the front wheel to turn the control arm has to be slightly modified. Moving the rear point forward could be an option, however, as the angle between the two arms decreases the forces in the arms increase very quickly. All of this means that in order to maintain a comfortable factor of safety for unknown loads, fatigue, heat affected zones, etc. I need to build a fairly strong set of lower control arms.

The steering rack has a home and mounting should not be too difficult. In order to comfortably clear the frame rail, it is shifted to the passenger side by 0.75in. This is achieved by lengthening the rack by 1 inch with a spacer which account for 0.5 inches of the shift and decreasing the rack travel by 0.5in which increases the turning radius slightly, but still results in an acceptable curb to curb of 36 feet. The rack is a RHD steering rack flipped upside down to account for the car being front steer. The tie rods are off a CX-5 with the standard Honda tie rod ends. Getting to a M14-1.5 rack makes things a lot easier.

The lower control arm is plated to increase the torsional stiffness and as it sits deflects 1mm at the ball joint under max braking. I may do further efforts to decrease this but in theory it is slightly below infinite fatigue life even with a max braking load case. The upper control arm is near identical to the rear with the only change of the orientation of the ball joints, a change I may make on the rear as well in the future. The upper will be used to control the camber as well as castor with slotted mounts to allow fore-aft movement, or multiple mounting holes and flip-able mounts to give 4 discrete locations.

The ARB is a fun one, due to balancing the location on the control arm to reduce the forces and decrease the bending load on the LCA with packaging and wheel clearances. I would like to use the old cup car rear splined bars as they are very easy to find on eBay and can be easily swapped out for different known stiffnesses as well as bought new. I am also somewhat certain the 28” ones are the same as sprint car torsion bars. It will likely be housed in an outer tube with bushings on each end and a healthy amount of grease inside. The end links will either be machined by PCBWay again, or I’ll buy something off the shelf.

A civic radiator is in place in the model which should provide enough cooling for light track use with a quality fan, but there is more room if I decide to go larger. As it will be a street car most of the time, the fan will be the most important part to staying cool.

Finally, I modeled a brake caliper bracket which means the corner assemblies are fully modeled. Not having a drivetrain at the front of the car does make this design process a lot faster.