LocostUSA.com

Hello everyone.

Ok, the introduction is out of the way...phew. I've been a pretty serious lurker for the past 6 months or so, I registered to make a post saying hi to a member here that I've spoken to in person, then shut up and kept reading and researching before deciding to finally make this post. I finally feel that I have sufficient parameters to work to and want to get down to all the details and make sure I'm on the right path with this project.

My plan is to build a mid-engine exo car. I know a lot of builders that go with this style car head in the atomesque direction but I plan to avoid that as much as possible. It will be reminiscent in the fact that it will use a transverse 4cyl in a mid engine configuration but my hope is to change the shape to be unique. From my research here and my initial planning it's turning out to be a bit of midlana, a bit of pook's mako and a bit of something else.

So lets get onto the details.

The donor will be a 2007 Mazdaspeed3. This is currently my daily driver and is very familiar to me. I know the car, I know the engine, I know the tuning. I've read the duratec/mzr thread and smiled a bit at just how different the disi is from the standard mzr engines and how much work will be involved. Why this engine? In truth it's because I've come to love it and it's little intricacies. As-is this engine would be a blast in this type of car, 350ft-lbs on tap in a chassis half the weight of it's current house. The transmission is bullet proof, the axles are stout, power can be as high as I want.

I'm getting a bit cathartic there, I really do love my car to death.

The purpose of this car will be high performance street with occasional track days/races as well. I don't plan to build to any particular rule book or regulations.

So onto the build.

The plan is to use the entire drivetrain as-is, just plunked into the back of the chassis. Rear track width will be the same as the current front track width of the ms3...or at least the hub face distance will be the same, track width will change based on wheels. Because of this I would like to keep the wheel base similar as well to keep stability at speed. The chassis will be designed as a 2-seater placing the occupants completely in the wheel base. If all goes according to plan the seats will be rigidly mounted (fixed back racing seats) with (potentially) movable pedals and steering to accommodate alternate drivers. Fuel tank will be between the seats and engine with a dual bulkhead setup for safety. Chassis width will be determined by chosen seats and center console necessity width. I will also try to keep a decent width in the footwell (something most locosts don't do well IMO) for comfort. Roll bar will be integral and quite high by many standards to clear the broomstick test...I'll get to the reasons for that in a bit.

Suspension is where my interest has been piqued lately as it was my area of least knowledge for the longest time so I'll try to keep this concise. Double wishbones on all 4 corners, inboard pushrod coilovers with 1:1 motion ratio, 4" bump, 2" droop, adjustable sway bars if I get that far. Geometry is up in the air a bit, I need to figure out uprights and will be limited in the rear by engine compartment size which will pretty much figure out a-arm lengths on its own. Steering rack will most likely be from a miata with extended tie rod ends; for multiple reasons I don't want to use the ms3 rack but it's not been completely ruled out. Uprights are undecided, I may use miata units (I already have many miata parts) on the fronts or make my own, not sure yet. I'm leaning more toward my own so I can run 100% Ackermann and drop the lower ball joint height at the same time. Rear uprights are completely unplanned as of now.

Onto a few numbers:

5" ride height (chassis bottom to ground). Keeps chassis from scraping at full bump.
65" hub-to-hub width in the rear. This keeps the axles "stock". I need to take final measurements but this should be quite close
104" wheel base, this is the same as a stock ms3.
48" roll bar height (at the top). Explanation to follow.
Approximately 148" overall length, this is really unimportant now it's just a figure I've come to through some sketches
15-20 gallon fuel tank for some semblance of range...this can be a very thirsty engine

For the suspension...

100% Ackermann
1* camber gain for first inch of bump
1* camber loss for first 2* of roll
As a side note the previous 2 were numbers I came up with through trial and error with vsusp over the past month or so. It was good affirmation when I saw that cheapracer here actually designed around the same numbers.
5* static caster
1.5-2*/1-1.5* static camber front/rear
1-1.5" front roll center height with minimal y-axis change through bump or roll
1-1.5" roll axis inclination front to rear (rear higher than front)

I think if I can hit those numbers then I can make this car handle. The chassis should be tunable at that point, at least good enough for a home built car.

Overall weight will hopefully be in the 1400-1500lb range, not unobtainable with careful design but also not a given.

When this design comes together it may be the biggest locost-type build on this forum and I'm perfectly okay with that. I'm 6'3 and mostly torso. I have quite wide shoulders and hips which gives me a very wide stance and my long torso means that while sitting my head is at the same height as someone 6'7. With a helmet on I want to clear the broomstick test so the roll bar will be high. Since I'm designing this as a trackable road car I plan to put the most comfortable racing type seats I can in it which are looking like corbeau fx1 wides to accommodate larger passengers (i.e. many of my friends) which will widen the cockpit as well. As long as I keep it in the track I'll be happy and initial measurements are putting me well within that range.

As far as myself I'm a 24 year old mechanical engineer by trade. I have a pretty good working knowledge of (and constant access to) Solidworks and will do my design there once my points are hashed out. I modeled a preliminary chassis to get a feel for it and the results were okay but I found a myriad of issues to work around and I've shifted my focus to laying out the suspension before finalizing the chassis to avoid unnecessary tubes and keep load paths well oriented.

I've done a number of measurements around my car to properly lay out the engine compartment and suspension which will come up when I can post up a chassis.

That all felt like a lot of rambling but I think I got all of the pertinents out there. Any comments or criticisms will be well received and I hope to justify any of my decisions where necessary. Over the long haul I will be posting up designs and discussing ideas here. If any of my goals seem lofty don't hesitate to tell me, I'm still learning a lot of the details (been doing my reading, especially on suspensions) and have done my best with what I already know.

I'm looking forward to furthering the project and couldn't be in a better place for it.

_________________
im CAD hat's passt

To add a bit of visual here is a link to my vsusp model showing my suspension layout. This is using custom defined uprights which I believe I will have to go with to get correct geometry. Vertical RC change is minimal, I have my downward sloping roll axis, camber changes are within my realm and I have enough travel to meet my original specs. I may change the specs later as well because I found that with this setup I can lower to a 4" ride height all around and keep both RC control, the correct roll axis, and the camber curves if I have enough UCA adjustment to lengthen .25" to correct the static camber gain. Overall I'm very happy with it and I will begin modeling the chassis using these numbers.

You may notice as well that I plan to use 17" wheels in the build, and that's purely because that's what I already have on the car. I will plan to use the shown 245/45/17 tires (currently on the ms3, albeit with 17x9 +48 wheels) and have an idea to use the 0-offset wheels as well, same width.

Hopefully in the next few days (maybe by the end of the weekend?) I'll have some sketches to post up of the chassis, or at lest a basic layout with enough planes and axes to define the chassis.

_________________
im CAD hat's passt

I laugh a little when I see that I thought I could pull together a fairly functional model in a weekend....

In case anyone cares I'm making progress much more slowly than I anticipated but I'm definitely moving forward. I've changed my concept a bit and took some more measurements and have suspension pretty well figured out, right now the challenge is making the frame rails look like I want them to and function like they must. It's slow going but with my hashed out figures it should pick up steam.

_________________
im CAD hat's passt

Forward is good! Everything seems to take longer then you think or plan for. Just keep at it.

_________________
Marcus Barrow - Car9 an open design community supported sports car for home builders!
SketchUp collection for LocostUSA: "Dream it, Build it, Drive it!"
Car9 Roadster information - models, drawings, resources etc.

I've been doing a good bit of modeling work trying to figure out the best way to really nail it down as a model. So far I've learned a whole lot of ways to work around the problems I seem to come by and what I really need to do is work in weldments...which I haven't been doing. In the future there will be a lot of work redoing the models I've already done in various ways until I'm really happy with it but for now here are some basic tube layouts. There isn't a floor yet, the engine box is not quite finished and there is no scuttle/console or front triangulation. At the very least it now has a shape though.

_________________
im CAD hat's passt

Any reason this tube needs to be bent like this? If you had it run lower and parallel to the ground it would provide better protection and eliminate the need for the "basket" to hold the seats.

_________________
-Andrew
Build Log
Youtube

No particular reason it's just aesthetic. I know it will be difficult to fabricate (that lower piece is actually 4 separate tubes) but I'm not a fan of the looks of a chassis with basically a floor plane and the curved top tube. The MEV rocket is setup that way and I don't like it.

The floor basket is okay with me and i plan to run a plexi sheet over the lower of the cockpit to limit exposure and whatnot.

_________________
im CAD hat's passt

Looks great!

I'm in a very similar design phase with my project, learning to work with weldments and 3D sketches, despite using Solidworks for years. Also, with your suspension geometry design, I saw that you used Vsusp... how did you determine what values you wanted for your car? I understand what all the different parameters mean, and what the average values should look like, but how do you know how it will translate to your car's handling characteristics? More to the point, did you base your geometry off of someone's proven design, or did you use theoretical numbers? I ask because that's something I'm struggling with at the moment.

I love the shape of the chassis you're designing. Do you plan to bend the tubes yourself or have them CNC bent? I'm working with square tube simply for cost reduction and ease of manufacture, at the expense of ascetics.

One last thought, you mentioned using plexi sheet for some bodywork, which is something I've been considering... but using it throughout the chassis in lieu of aluminum bodywork. Thoughts?

Anyway, looks great so far, keep us updated!

_________________
Check my Wordpress blog >>> Project: Garagisti Alpha
my LocostUSA build thread >>> Garagisti Alpha: Exo Mid Engine Design and Build

The suspension values are all based on theoretical "good handling". I've laid out the miata suspension before but I knew my chassis would be vastly larger, wider, and have a completely different weight distribution so I didn't read too much into it. There's also the fact that all production cars, not matter how well they handle, are designed to fit a specific chassis package which doesn't necessarily apply here.

The parameters I listed in the OP were the best I could gather from as much reading as I've been able to do. My general requirements/parameters were:

low roll centers with a downward sloping inclination toward the front
zero positive camber in any circumstance I could manage
ground clearance for the chassis under full bump
general balance of camber loss in roll vs. camber gain in bump

On top of that I had to work around my engine package which is, unfortunately, quite wide. The package is stout and powerful though so I was willing to compromise.

The realization I came to with suspension is that it cannot be perfect without active suspension and I don't have the time or money to develop that system....or buy it from Williams. So what I did instead was use general guidelines and create numbers that hit the spectrum. The rest will come down to chosen shocks and springs (TBD when I have a more accurate idea of weight and distribution) and roll bars that I will probably eventually fabricate.

For chassis fabrication I plan to do as much of it by myself as I can. I'm not making that easy by creating lots of curves but since this is an exo car the aesthetics are important (to me) and I'm not willing to sacrifice looks for ease of manufacture. I will buy a tube roller from HF and upgrade it as necessary. I'm not sure about the bends but they're currently all spec'd to the JD squared 5.5" radius for 1.5" OD .083 tubing so if it works out in my favor financially I will take that route. Or I'll beg someone here with a roller to do the work for a small fee. I don't plan to pay to have them done professionally but that may change in time.

I don't plan to do plexi for all the body work, it's just not flexible enough for a few things I plan to do. I may even go with fiberglass and create bucks for some special work but that's really getting ahead of myself. I also don't plan to use aluminum as body sheeting, it will be used for front and rear bulkheads as well as the undertray(s) but nothing else. Most of the car will be open ala the atomic one.

Funny you should mention weldments in Solidworks, the model I posted shots of was done entirely using sweeps and extrusions. I made it quite far with that model but have restarted today using weldments and it's sooooooo much easier and faster. And my model tree looks perfect now. I will post up a more complete chassis later today, one with a floor and scuttle and whatnot. It's really taking shape.

lastly thank you for the compliment (or at least I'll take it as such) on the looks. I have about 100 sketches of the chassis in various dimensions and was never really happy with them structurally. I'm absolutely thrilled about this one though and it feels good to be working toward a "final" product in terms of the model.

I just bookmarked your blog to read in full at a later time as well

_________________
im CAD hat's passt

work is going well.

I managed to remake the entire chassis again today, this time using all weldments. Looks largely the same but is more complete. Needs a center console and there will be verticals at the suspension points as well as a bolt-on larger roll bar brace ala Pook's Mako. If you can't tell by my design I quite like the Mako.

More to come soon.

_________________
im CAD hat's passt

looks very nice i think if time and money werent as big a problem for me id opt for a frame with more complexity like yours. i really like the curves you have going. question tho, as some one who works in solidworks daily i use a lot of weldments and the occasional 3d sketch in them, how many 3d sketches did it take to make that? i only ask bc 3d sketches and curves are not necessarily very good friends in my experience. i suppose it all depends on how many construction lines you add

_________________
Anyone who has never made a mistake has never tried anything new. ~Einstein

Build log http://locostusa.com/forums/viewtopic.php?f=18&t=15177

Wow, chassis looks great! I definitely see some of pook's design there, especially in the front end. It's funny you mention that car, that was the one that made me actually start planning to build one of these, it's the best looking exo-car around, IMHO. I know what you mean about not being pleased with sketches, I could not copy that Mako shape to save my life. Anyhow, great progress... looking forward to updates!

_________________
Check my Wordpress blog >>> Project: Garagisti Alpha
my LocostUSA build thread >>> Garagisti Alpha: Exo Mid Engine Design and Build

See the pic below for my current model tree of 3D sketches. I could honestly do it in 1 but my method would make the page so cluttered I'd never find anything.

The way I've been doing it may be unique to me but it works. I only model half the car, not all of it. Once I sketch a major component or even just a standalone part I create the member and do any necessary trims. I then name the sketch and member and toss the trims into a folder to clean up my model tree for later use. Once I have an entire chassis half complete I mirror it about the longitudinal centerline to complete it.

Honestly curves are pretty easy, or at least I've figured out a way to do it. Every tube in that model is viable for me to make on my own if you know where the cut lines are...I plan to do sleeved butt joints between the more complex curves then grind flat to create a smooth finish.

Most of the models are laid out by first defining points in space and dimensioning them off of the 3 main axes. For my "upper main rail sketch" I defined 5 points. One is on the longitudinal plane, one is a straight line off of longitudinal (x-direction), then a series of 3 to make the curve using changes in all 3 coordinates. Point 3 is only dimensioned in the z-direction. Points 1, 2, and 3 define the bend in the nose and straight lines are drawn between them. A 3-point arc is defined using points 3, 4, and 5 and the arc is set coincident to each point. The kicker is to then define the arc as tangent to the second straight line from the nose, then radius the nose curve in the sketch. When you make a structural member out of these points it all looks nice and flows in a solid curve. If you want pics progressing through this let me know, I'll be happy to post them up

The real trick is knowing where to set all the points. I have a bunch of sketches with dimensions in them and in the latest model I have sketched axes to define my suspension pivot points so I can properly avoid putting tubes through them. I know the dimensions of a seat I plan to use as well to make enough space in the cockpit, I know my own dimensions to account for roll bar height and cockpit length for pedals, and I know the frame rail inner width of my donor car to fit the engine. Realistically the rest can be figured out later. I've also planned the build around 1.5" .083 and 1.0" .065 tubing so I can mimic realistic bends based on a JD squared bender with appropriate dies.

_________________
im CAD hat's passt

It could be worse.... could be raining.... lol. nice, Ive got mine going in 3 separate sketches. One for the passenger compartment one for engine and rear suspension, and one for the front suspension. Though admittedly mine is far simpler, mostly for my own ease of getting the first one on the road. i have a small collection of more complex and complicated drawings that will never likely be made simple due to fabrication complexity and lack of some simple/basic principles Ive managed to ignore or be ignorant of.

Def like where your heading Cant wait to see tubes being cut.

_________________
Anyone who has never made a mistake has never tried anything new. ~Einstein

Build log http://locostusa.com/forums/viewtopic.php?f=18&t=15177

Doing it in a single weldment with multiple sketches is the best option for me IMO, it keeps it all in one place but breaks it down so that I can edit a part without everything getting too muddled.

Before I cut anything I will model the entire suspension, wheels, brakes (hopefully)....everything. It's slow going at times but a proper model will make it all easier in the end. I'm hoping to do it right once and not have to make too many parts over again

_________________
im CAD hat's passt