LocostUSA.com

Dave H

I would not be overly concerned about weld penetration on the ID of the taper. You are going to have 80% of the parent metal holding the stripper bolt and just one small weld bead around the inside in one area.

I would be concerned if you are going with a welded on steering arm or a section of the arm welded on to the spindle. Plus you will NEVER weld the parts on in identical locations from left to right hand spindle. The steering attachment- holes are machined within a few thousands of an inch on all three planes.

You do not want the steering different from Left to Right hand side, Dave W

I had some fun in the shop tonight, learned a few things, and got a good part of the deal.

In order to invert my tie rod ends to the top of the Miata steering arm, I decided to make a tapered sleeve and bore a taper from above.

I made the outside taper on the lathe using the compound. I had a hard time trying to set the angle several times, without a good fit. Then I figured out to mount the tapered reamer in the tailpiece and use a 3-2-1 block to align the compound to the reamer. Worked perfectly the first time. (So much to learn!)

I pressed the fitting into the steering arm. I planned on pulling it back out for a photo shoot, but it stuck so well I don't think I *could* get it out! It stuck so well, I decided that welding it in place would not be needed. This proved almost true.

This is what it looked like in the arm:

(to be continued shortly...)

_________________
...nowadays people are so intellectually lazy and lethargic that they can't build ANYTHING with their hands. They'll spend hours watching whiny people marooned on an island, but won't spend a second adding anything to the world. -weconway
Visit my [Locost 7 build log]

So then I chucked it up in the mill to bore the taper. This was intimidating, as there is nothing perpendicular or true about the steering arm, so I expected a major battle. I was happily disappointed!

Just for laughs, I chucked it up in the vice on some parallel bars, with plenty of room below for the taper to pass through. This was clamping on the round sides of the steering arm end, which are rough cast and have an irregular profile any way you look at it. Plus, it was supporting the entire weight of the spindle. I really expected it to launch the spindle at my belly the instant the reamer started cutting. I was tempted to machine a flat on each side to clamp on, but it seemed fairly solid even with this marginal clamp.

I decided to gingerly start boring at 250 RPM and see what happened, prepared for the worst. Much to my surprise, it just pushed right through the cast iron like warm butter. I've been doing so much work with steel since I put my machine shop together, I'd forgotten just how easy cast iron works.

My goal was to bore slightly deeper than the original taper, so that the tie rod would contact about 2/3 of the parent material from above. Whatever was left of the taper would fill in the remainder, but likely not be critical.

Well, just over half way through, the insert came out. I decided I didn't care and pressed on to 2/3, as indicated by a Sharpie mark on the reamer.

I'm very pleased with the result. The first picture shows that the tie rod did not fully seat in the stock hole and does fully seat in the new hole. It doesn't show in the photos, but the small end of the taper does not reach the bottom of the hole in the stock config. Actually, you can infer that: since you only see threads emerging in the "after" photo, so you know there's at least as much thread inside the outer taper as there is inner taper showing above the casting...about 1/8". (Does that sentence make your head hurt too?)

So the bottom line is, I feel there is as much or more mating surface in this configuration than the stock config. Even if my clever insert had worked, I don't think it would have amounted to much. After doing all this, I do not recommend filling the taper with an insert or welding. Just ream it, put a hard washer under the nut, and run with it...especially for a car 1/2 the weight of the donor.

I also don't think you'd need a mill to do a good job. The alignment is not critical, so I think you could even get away with a slow hand drill. In fact, if you don't have a good vice on your drill press, it might be easier to just leave the spindle mounted on the car and use a hand drill than to try to jig it up in the drill press. Keep your RPM's low. There is no reason for anyone to be intimidated about doing this...it really is a piece of cake.

Well, that's it for tonight. I spent a couple of hours to do the first spindle. The second one will take all of 10 minutes...if I don't hurry! Next I'll dial in exactly where the rack needs to go with the new tie rod location. (Yeah, yeah...I know I should have analyzed that first, but it will be SOO much better than before, I haven't got much to lose, and the eyeball says this will be close to on-target anyway.)

-dave

_________________
...nowadays people are so intellectually lazy and lethargic that they can't build ANYTHING with their hands. They'll spend hours watching whiny people marooned on an island, but won't spend a second adding anything to the world. -weconway
Visit my [Locost 7 build log]

ps. That's a '93 Mazda 323 tie rod end, and it has plenty of angle for the job. Same threads/taper as the Miata ends, without the wonky bend. I haven't considered yet if the length is different.

-dave

_________________
...nowadays people are so intellectually lazy and lethargic that they can't build ANYTHING with their hands. They'll spend hours watching whiny people marooned on an island, but won't spend a second adding anything to the world. -weconway
Visit my [Locost 7 build log]

That's a great writeup, Hempy... I could use the same thing. Think you could turn a second set of inserts? A lathe/mill is the next tool I need on my list...

Dave, Nice work on the tapered insert. If you were to tack weld the insert on the lower surface of the arm, then ream from the top, then gringd away the weld, you should be able to keep the insert in place during reaming. That is if you still think you need it. Of course, you probably thought of that already.

When I did mine I just put my reamer into my 1/2" drill and drilled the hole freehand as a low speed, Mounted right on the car, too easy.
wayne

I guess I didn't state my summary clearly enough.

I don't think the insert is useful, and don't recommend it. My car won't have them.

Like El Wayne did, I think you could do just as well with a slow hand drill and the spindle on the car.

-dave

ps. But yes, I'd be happy to make some inserts if you want them, Nate. It's no trouble at all. Send me a PM with your address and I'll get them out Monday. After the fact, I considered leaving a large shoulder/flange for the nut to land on and compress the insert into the bottom taper, but then you start having to think about the insert and the stud converging and whatnot. I still think the insert is not useful in this application.

_________________
...nowadays people are so intellectually lazy and lethargic that they can't build ANYTHING with their hands. They'll spend hours watching whiny people marooned on an island, but won't spend a second adding anything to the world. -weconway
Visit my [Locost 7 build log]

I've gotten a few PM's, and it seems I've confused more than I've clarified. I guess I'm not explaining it very well.

How it ended up is that I've got tapers from both sides now. If you viewed it in cross-section, it would look kinda like a very slender hourglass. (See below) I can still install the tie rod from either side, and it seats soundly. (Although inserting from below now has almost *no* mating surface area, and I'd never actually install/drive it that way!)

If I had bored from the top to the same size as the factory taper below, the "waist" of the hourglass would be half-way through the arm, and both sides would have just as much mating surface.

I actually bored about 75% deeper from the top, so the waist is 75% down from the top of the arm. The new major OD at the top is just a few thou larger than the stock major OD at the bottom. Of the two "halves" of the hourglass, the new, upper taper has much more mating surface. The old, lower taper is only 25% of it's original height.

FWIW, the taper on the tie rod stud is not as tall as the steering arm's thickness. I really don't think it even reaches last 10-20% of the taper at the bottom of the hole anyway. It probably wouldn't even touch the insert much, if at all. That's why I don't think the insert is relevant.

Hope this is more clear...sorry for the confusion.

_________________
...nowadays people are so intellectually lazy and lethargic that they can't build ANYTHING with their hands. They'll spend hours watching whiny people marooned on an island, but won't spend a second adding anything to the world. -weconway
Visit my [Locost 7 build log]

This is what it looks like at ride height with the inverted tie rod ends. MUCH better than before, and nothing wacky needed to get the rods and the control arms to get along.

For photos at droop, ride height, +1.5" bump, +3" bump, and +4" bump, see Page 5 of this thread: http://locostusa.com/forums/viewtopic.p ... 9&start=60

Or see it in video here: http://www.youtube.com/watch?v=ubNMWZF67qU

_________________
...nowadays people are so intellectually lazy and lethargic that they can't build ANYTHING with their hands. They'll spend hours watching whiny people marooned on an island, but won't spend a second adding anything to the world. -weconway
Visit my [Locost 7 build log]

Good job Dave. Just to ease any concern you might have on those double tapers, its a common practice on Ford hot rods to do exactly what you've done. The tie rods and drag link holes in the steering arms are reamed from the other side and the tie rod end can be inserted from the bottom, lowering the steering linkage to match a dropped axle.

Hot rodders have been doing this since the thirties and its a time tested approach. My own 32 roadster has the same double taper on the 1940 steering arms which were reamed by some unknown rodder in the distant past.

_________________
Jack Collins

Kind of an old post to comment on, but I did the above for my rear suspension - and won't do it again.

The problem isn't in bump, it's in droop when the suspension returns to its neutral position. That nice toe-in on bump is now toe-out on droop, right? So mid-corner when the car hits a bump, the wheels will toe in more, then toe out as the suspension recovers. This is designed-in bumpsteer.

I'd instead just aim for zero toe change to avoid the whole issue.

_________________
Midlana book: Build this mid-engine Locost!, http://midlana.com/stuff/book/
Kimini book: Designing mid-engine cars using FWD drivetrains
Both available from https://www.lulu.com/

From what I've seen it's not uncommon to see some toe-out used on modern AutoX alignments to help with turn-in to bring the back end around. I have not seen toe-out used on real race tracks with a modern suspension on RWD coupe or sedan type cars.

If I wanted an easy to drive street layout then I would design the front end have the steering at maximum length (width?) at neutral ride height, and then have the tie-rods short enough that the car went into very mild toe-in when the suspension went into droop or bump, but only very slightly. Mild toe-in is easier to drive and more comfortable than toe-out. Going into a mild toe-out situation with the front end isn't all that bad as long as it's minor. I've driven cars with plenty of bump steer on both the track and street and it's controllable, but not comfortable. Too much bump steer can also be hard on tires, causing excessive wear. toe out will also make the car feel more "jittery" or "darty" while a tiny bit of toe in will help the car feel more settled, especially at higher speeds.

If you are designing for an IRS then make sure the back end doesn't go into toe-out under compression; that condition can lead to what alot of drivers call snap oversteer. As you turn into the corner and the car leans, the outside rear corner will toe-out and the car will turn even faster.

The driver should drive the car, not the rear suspension. The first case, mentioned by Kurt, is a cause of trailing throttle oversteer. Get your self good and scared, then take your foot off the gas or even go for the brake and now you get to cope with oversteer too! It is a very bad situation. I think it is the root cause of several 911's I have witnessed leaving the track in a little round ball shape.

_________________
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 set up the rear for very slight toe-in in both bump and droop. What I failed to think through was what happens on the way out of a turn.

Come into the turn on the brakes, rear goes into droop and toes in - good so far. Now get on the gas and the rear wheels start returning to straight ahead, which, coming from a toe'd in condition is the same as moving toward a toe-out condition. Then as the chassis is loaded up under acceleration the rear wheels go into bump where they toe in again.

In hindsight it's a case of overthinking a solution. So for me, no toe control anywhere, front or back.

_________________
Midlana book: Build this mid-engine Locost!, http://midlana.com/stuff/book/
Kimini book: Designing mid-engine cars using FWD drivetrains
Both available from https://www.lulu.com/