Wheel Alignment

[DanH]

Recently did a field repair involving ad hoc wheel alignment using whatever was handy. It led me to think about my own alignment, as it hadn't been checked since the fuselage was upside down in the shop.

Here's the simple, reliable, repeatable, hardware to do it. The bits are cut from scrap aluminum 3/4" thick. The big block is just a spacer. It ensures the gauge is referencing the wheel, not the relatively squishy tire. The little clamp block is drilled through 0.250" for an AN4A, and tapped through for a 1/4-20. Bolt it to the standard Vans nut, use the 1/4-20 to tighten on an 8 ft length of 1-1/2" aluminum angle from the rack in big box home store.



Ignore the holes in the spacer blocks, an artifact of a previous use.



Stretch a centerline string and measure to it, between the gear and at the ends of the 8 ft angles.



So what was learned? Tail down, the right gear had 0.671 degree toe in. The left was 0.075 degree in.

Because of positive camber, toe in is increased as the tail is lowered. There will be some camber variation between airplanes. My Grove gear has quite a lot of positive camber, so the difference is 3/16" in four feet, per wheel. It's more than enough to make tail down somewhat squirrely compared to tail up.



Grove sells nice tapered shims in 1/4 degree increments. Using them would not result in perfect alignment, as 0.071 and 0.671 are not divisible by 0.250. I elected to install strap shims. A bit of CAD modeling allowed me to determine the exact shim thickness required, 0.0018" and 0.0161". The material is readily available. Just pick up a set of cheap feeler gauges and make shims from the appropriate strips.



Plenty of debate about the merits of toe in and toe out. I'm not going there, except to say do as you please. My choice was to align for exactly parallel with the tailwheel on the ground and the weight of the aircraft on the gear. And I got it too, within the thickness of the mason string. It means I'm toed out slightly at longerons level and no load. In that condition the aircraft is still pretty much flying, and not so much rolling, so I don't think the toe out matters. Did a few landings this AM, and liked it. Some crosswind, so it's hard to quantify a subjective observation just yet.

Anyway, regardless of your preference, I think you'll like the parallel gauge setup a lot.

 

[scard]

Nice! Thanks for sharing.

 

[swjohnsey]

When doing wheel alignment on motorcycles I would use 8' florescent light bulbs held against the side of the wheel with rubber bands. Positive camber usually requires toe-in for straight track. Changing from toe-in to toe-out during wheel travel to be avoided. Causes bump steer in cars, not sure what it would be called in an airplane. Also something call toe-plates that racers use to set toe, pretty cheap using a couple of tape measures. Third option is something called Gunson Trakrite. It measures scrub/toe directly by rolling one tire over it. My favorite.

 

[Scott Hersha]

I tried a similar setup for checking toe in/out, but probably not nearly an accurate as your setup. I like your setup, but not sure what I can do about it with my round gear RV6 (same with RV7, 9, RV4). Any e there’s a way of shimming the round gear birds, but I haven’t figured it out yet.

 

[fl-mike]

Recently did the same. I use HF 6' levels as straight edges. Found they are about the same cost as buying angle. I was surprised to see how close the "tire" method was to taking the measurements off the legs. I have a 3D "hat" designed to screw on the wheels using the hubcap holes that holds the levels, but our 3D printer is acting up, so I pulled the axles and clamped to the gear legs. Like you, I have Grove legs and quite a bit of camber.
I went for zero at level/weight-off, and have toe-in with the tail down. I was initially over the spec on toe in (level), and shimmed to zero. Seems better behaved now and I'm feeling better about my landing skills! (Although aligned during the build, I had been struggling with "twitchiness" on landing, especially when the tail came down, so thought I would re-check.)

 

[JonJay]

Dan - you’ve been flying for some time. Did you notice any bad behaviors, unusual tire wear, anything? Just curious why you would care to make a change at this point.
Not taking away from your endless and quality contributions, and that is reason enough, just curious.

 

[scsmith]

I shimmed out the camber as well as the toe-out. nearly zero in both longerons-level and in 3-pt attitude.

 

[DanH]

Dan - you’ve been flying for some time. Did you notice any bad behaviors, unusual tire wear, anything? Just curious why you would care to make a change at this point.
Not taking away from your endless and quality contributions, and that is reason enough, just curious.

It was wearing the outside of the tires, and was always a bit twitchy when the tail came down, on asphalt. Then another -8 owner asked me about the geometry, then came the field repair alignment check, then came a 1750 mile one day run for home. Not much to do but sit there and think about it.

Perfect alignment may not really improve anything, but I like the idea. Gimme a few dozen landings and we'll know more.

 

[Taltruda]

It was wearing the outside of the tires, and was always a bit twitchy when the tail came down, on asphalt. Then another -8 owner asked me about the geometry, then came the field repair alignment check, then came a 1750 mile one day run for home. Not much to do but sit there and think about it.

Perfect alignment may not really improve anything, but I like the idea. Gimme a few dozen landings and we'll know mor

 

[Taltruda]

Dan, you mention excessive positive camber.. do you think Grove could or would put a little more bend in the gear for you? It may take them to anneal, bend, then re heat treat, but it may be worth it to get rid of the excessive camber?

 

[DanH]

Dan, you mention excessive positive camber.. do you think Grove could or would put a little more bend in the gear for you? It may take them to anneal, bend, then re heat treat, but it may be worth it to get rid of the excessive camber?

A little less bend. I doubt Grove would be interested.

The bends in the stock steel leg are 51 degrees upper and 46 lower. Note the total is 97 degrees, meaning when unloaded, the stock leg has 7 degrees of camber. I don't know if the Grove legs are the same.

Camber is reduced when the gear is loaded. Final camber depends on load at that moment. It's possible to shim an RV-8 axle to zero camber, but only at one given weight.

 

[chaskuss]

Dan those shims are available through automotive vendors. I know this because I had to change the alignment on one of the rear wheels of my car some years ago. My car is front wheel drive. I'm not at my shop at the moment but when I get back I'll try and find the box and get the manufacturer and part numbers. As I remember there were five or 10 shims to a box. I believe they came in 1/16th & 1/ 8-in thicknesses.

 

[DanH]

Dan those shims are available through automotive vendors. I know this because I had to change the alignment on one of the rear wheels of my car some years ago. My car is front wheel drive. I'm not at my shop at the moment but when I get back I'll try and find the box and get the manufacturer and part numbers. As I remember there were five or 10 shims to a box. I believe they came in 1/16th & 1/ 8-in thicknesses.

Probably too thick, but a fine idea. In my case a big correction only required a 0.016" shim.

 

[Flyguysr5]

Recently did a field repair involving ad hoc wheel alignment using whatever was handy. It led me to think about my own alignment, as it hadn't been checked since the fuselage was upside down in the shop.

Here's the simple, reliable, repeatable, hardware to do it. The bits are cut from scrap aluminum 3/4" thick. The big block is just a spacer. It ensures the gauge is referencing the wheel, not the relatively squishy tire. The little clamp block is drilled through 0.250" for an AN4A, and tapped through for a 1/4-20. Bolt it to the standard Vans nut, use the 1/4-20 to tighten on an 8 ft length of 1-1/2" aluminum angle from the rack in big box home store.

View attachment 65519

Ignore the holes in the spacer blocks, an artifact of a previous use.

View attachment 65520

Stretch a centerline string and measure to it, between the gear and at the ends of the 8 ft angles.

View attachment 65521

So what was learned? Tail down, the right gear had 0.671 degree toe in. The left was 0.075 degree in.

Because of positive camber, toe in is increased as the tail is lowered. There will be some camber variation between airplanes. My Grove gear has quite a lot of positive camber, so the difference is 3/16" in four feet, per wheel. It's more than enough to make tail down somewhat squirrely compared to tail up.

View attachment 65524

Grove sells nice tapered shims in 1/4 degree increments. Using them would not result in perfect alignment, as 0.071 and 0.671 are not divisible by 0.250. I elected to install strap shims. A bit of CAD modeling allowed me to determine the exact shim thickness required, 0.0018" and 0.0161". The material is readily available. Just pick up a set of cheap feeler gauges and make shims from the appropriate strips.

View attachment 65526

Plenty of debate about the merits of toe in and toe out. I'm not going there, except to say do as you please. My choice was to align for exactly parallel with the tailwheel on the ground and the weight of the aircraft on the gear. And I got it too, within the thickness of the mason string. It means I'm toed out slightly at longerons level and no load. In that condition the aircraft is still pretty much flying, and not so much rolling, so I don't think the toe out matters. Did a few landings this AM, and liked it. Some crosswind, so it's hard to quantify a subjective observation just yet.

Anyway, regardless of your preference, I think you'll like the parallel gauge setup a lot.

Great idea on bolting the plates on to hold the straight edges! Question....With out the wheels sitting on grease plates how are you getting them to go to natural position from rolling forward or backward or lifting off the ground etc?

 

[chaskuss]

Probably too thick, but a fine idea. In my case a big correction only required a 0.016" shim.

I just dug up the shims. Moog part number K977-1 are .010" thick, 15 to a box. Moog K977-2 are .020" thick, also 15 to a box.


My apologies about the inverted photo. I'm doing this on my phone so I can't flip it around.

 

[DanH]

Question....With out the wheels sitting on grease plates how are you getting them to go to natural position from rolling forward or backward or lifting off the ground etc?

Grease plates are fine, but here at least one wheel has been rolled into place on an old short nap carpet.

 

[Flyguysr5]

Grease plates are fine, but here at least one wheel has been rolled into place on an old short nap carpet.

Short nap carpet. I knew you probably had a trick I had not heard of. Thats why I asked. Thanks, I always enjoy your tips.

 

[vfrazier]

One clue that your wheels need better alignment, aside from tire wear, is if the tires squeak continuously when rolling the aircraft across a smooth floor. Continually squeaking tires when rolling on a shiny, painted concrete floor can be a sure sign of misalignment.

Having sold a LOT of tailwheels over the years allowed me to talk about a LOT of RVs. By far, the RV-8s are ones the most commonly mentioned as "twitchy". (Note: I am NOT bashing the plane. It's a great airplane!) I have often wondered how many RV builders actually checked their main gear wheel alignment. I suspect that many never give it a thought unless their birds are eating up the tires.

So there may be room for improvement on many RVs, although changing the alignment on RV-4s, 6s, and 7s, etc. with the bent rod gear is problematic for sure. Changing the alignment on an RV-8 should be relatively easy, so why not check it?

IMHO, setting the wheel alignment with the wheels off of the ground is pointless. The alignment will likely be different when the rubber hits the asphalt.

We recommend using greased plates for setting the alignment on the Rockets. This is so that the a/c can be raised and lowered without worrying about rolling it forward. How you check the alignment is irrelevant as long as you're measuring the wheels when they are in a normal rolling position.

It is probably a good idea to drop a plumb bob from your spinner and make a line to the tailwheel and check to see if your main wheels are reasonably parallel to this line. Yes, the engine thrust offset skews this line a bit, but that should be OK. Afterall, you don't want the engine pulling left while your wheels are pulling right.

Dan's technique is very good. However, if you're short on machined blocks, simply clamping a straight angle to the brake discs is a viable option too. Obviously YMMV here, but it works on the Rockets.

One other Rocket suggestion, that should be valid for the RVs too, is to load the a/c to an average flying weight. For the Rockets we recommend having about 1600# as a typical weight, i.e. the aircraft plus full fuel and a pilot. (I have been known to purchase sandbags, or similar, at a big box store to load the plane, then return them to the store afterwards. ) Use whatever weights are needed if you're doing this during construction!

It is important to note that for most of our airplanes there is no one perfect alignment, and no one perfect weight. And having the tail up or down is a debatable point too. Nonetheless, I'm confident enough to say that for most of us the best average alignment can be found at:

• your typical flying weight
• measured with the tail on the ground (unless you have an "A" model, of course)
• camber can affect your alignment, but it's usually negligible, so don't sweat it... usually.
  

 

[fl-mike]

...snip

It is probably a good idea to drop a plumb bob from your spinner and make a line to the tailwheel and check to see if your main wheels are reasonably parallel to this line. Yes, the engine thrust offset skews this line a bit, but that should be OK. Afterall, you don't want the engine pulling left while your wheels are pulling right.

Dan's technique is very good. However, if you're short on machined blocks, simply clamping a straight angle to the brake discs is a viable option too. Obviously YMMV here, but it works on the Rockets.

One other Rocket suggestion, that should be valid for the RVs too, is to load the a/c to an average flying weight. For the Rockets we recommend having about 1600# as a typical weight, i.e. the aircraft plus full fuel and a pilot. (I have been known to purchase sandbags, or similar, at a big box store to load the plane, then return them to the store afterwards. ) Use whatever weights are needed if you're doing this during construction!

It is important to note that for most of our airplanes there is no one perfect alignment, and no one perfect weight. And having the tail up or down is a debatable point too. Nonetheless, I'm confident enough to say that for most of us the best average alignment can be found at:

• your typical flying weight
• measured with the tail on the ground (unless you have an "A" model, of course)
• camber can affect your alignment, but it's usually negligible, so don't sweat it... usually.

On the 8, there is a handy rivet on the exhaust tunnel on the centerline for dropping a plumb bob. Just make a notch in some scrap sheet to position the string, and tape the sheet to the tunnel with the notch over the rivet.
Van's has you set and check alignment level, unloaded, so that is the standard practice. (I can kindof see an argument for checking loaded and tail low though! Especially if you're a 3-pointer))
I disagree on the camber part. There is a large change in toe-in on a taildragger if you have significant camber.
Camber may not be as pronounced on the 4/6/7/rocket tapered rod. The 8, and my Grove gear, has some decent camber. I measured an increase in toe-in of about 0.5-0.6 degree per side tail up to tail down with weight on the wheels.

Tail Up: (not actual Grove gear! )



Tail down:

 

[DanH]

Nice illustration Mike. My conclusion also.

Just made a trip to Baltimore and back with several stops for gas and friends. I am delighted with the new setup, to wit, dead parallel with the tail down. The resulting small toe out with the tail up doesn't generate anything noticeable, but there is a real difference with the tailwheel rolling, on the runway or the taxiway.

 

[DanH]

IMHO, setting the wheel alignment with the wheels off of the ground is pointless.

Yep. That is for wheel pant alignment.

It is probably a good idea to drop a plumb bob from your spinner and make a line to the tailwheel and check to see if your main wheels are reasonably parallel to this line. Yes, the engine thrust offset skews this line a bit, but that should be OK. Afterall, you don't want the engine pulling left while your wheels are pulling right.

Did some fooling around with different methods of establishing centerline. The trouble with a plumb bob dangling from anything is the requirement for the aircraft to be perfectly level, side to side. Eventually I concluded reference to the fuselage isn't really necessary. The goal is get the mains parallel. Simply measuring across between the gear legs to locate a center was fine. Tie a string around the tailwheel stinger, stretch it to bisect the measurement across the gear, and measure from the string to each end of the parallel bars.

...simply clamping a straight angle to the brake discs is a viable option too.

Excellent idea!

One other Rocket suggestion, that should be valid for the RVs too, is to load the a/c to an average flying weight.

Loading appears to be very important with swept round legs, but much less so for the unswept slab gear.

 

[Mikeyb]

I held 3/4” straight steel angle against the brake disk with tapered framing shims against the gear leg. It’s surprisingly robust and repeatable. These parallels are 36” long and requires estimating the tape to 1/64” to get 0.05 degree accuracy.

 

[DanH]

These parallels are 36” long and requires estimating the tape to 1/64” to get 0.05 degree accuracy.

Therein we find the joy of 8 foot parallels

 

[HFS]

Just For Clarification ...

For accuracy of our calculations for toe in/out, we probably would like to be able to see (reliably) a .1* variation.

Using a 36" spanner (18" to CL), a .1* variation is .03141"; and, using an 96" (48" to CL) spanner, a .1* variation is .08378" - a much easier (x 2.7) measurement when using hand held tape measures.

Even though the process is the same, the "resolution" may suffer a bit ...

HFS

 

[DanH]

Even though the process is the same, the "resolution" may suffer a bit ...

Resolution can be further improved by ditching the tape measure. I found a 48" hard ruler, blocked up to just under the string, was easier to use with accuracy. No sag, no hands, etc. See the first post.

Ever taxi an airplane and heard a rattling brake caliper? Typically it's because the disk is warped just a bit. Probably don't want to clamp the parallels to the disks if yours is a rattler.

Knowing Dave, he probably went straight to trig on a calculator. CAD works for lazy types like me, and can determine the exact shim thickness required. And Grove has a nice calculator on their website. See the shim page.

It's interesting to see how very small misalignment results in fairly large steering inputs. Assume the pilot sets the tail down at 45 knots, which is 76 feet per second. In one second (say "one thousand one" as you read this) the 0.671 degree toe in I found in my right gear steers the airplane 10.6 inches to the left. A full degree would be about 16 inches. Dave's suggested "less than 0.1 degrees" would thus be less than 1.6 inches.

Got thinking about the many differences between the -8's straight slab gear and the swept gear on other models. Here's a top view:



Given a swerve generates a slip angle (here drawn as 5 degrees) and thus a side force, the swept leg bends inboard, restoring a tire alignment closer to straight ahead. The bend won't yield the full five degrees, but it will be close, in particular with long titanium (low modulus) legs.



I knew those Rocket guys were cheating

 

[Jjackh10]

On the 8, there is a handy rivet on the exhaust tunnel on the centerline for dropping a plumb bob. Just make a notch in some scrap sheet to position the string, and tape the sheet to the tunnel with the notch over the rivet.
Van's has you set and check alignment level, unloaded, so that is the standard practice. (I can kindof see an argument for checking loaded and tail low though! Especially if you're a 3-pointer))
I disagree on the camber part. There is a large change in toe-in on a taildragger if you have significant camber.
Camber may not be as pronounced on the 4/6/7/rocket tapered rod. The 8, and my Grove gear, has some decent camber. I measured an increase in toe-in of about 0.5-0.6 degree per side tail up to tail down with weight on the wheels.

Tail Up: (not actual Grove gear! )

View attachment 65708

Tail down:
View attachment 65709

Good illustration. To me it kinda makes the case to check it loaded. A bit of weight on the gear should reduce the camber.

 

[vfrazier]

Resolution can be further improved by ditching the tape measure. I found a 48" hard ruler, blocked up to just under the string, was easier to use with accuracy. No sag, no hands, etc. See the first post.

Ever taxi an airplane and heard a rattling brake caliper? Typically it's because the disk is warped just a bit. Probably don't want to clamp the parallels to the disks if yours is a rattler.

Knowing Dave, he probably went straight to trig on a calculator. CAD works for lazy types like me, and can determine the exact shim thickness required. And Grove has a nice calculator on their website. See the shim page.

It's interesting to see how very small misalignment results in fairly large steering inputs. Assume the pilot sets the tail down at 45 knots, which is 76 feet per second. In one second (say "one thousand one" as you read this) the 0.671 degree toe in I found in my right gear steers the airplane 10.6 inches to the left. A full degree would be about 16 inches. Dave's suggested "less than 0.1 degrees" would thus be less than 1.6 inches.

Got thinking about the many differences between the -8's straight slab gear and the swept gear on other models. Here's a top view:

View attachment 65739

Given a swerve generates a slip angle (here drawn as 5 degrees) and thus a side force, the swept leg bends inboard, restoring a tire alignment closer to straight ahead. The bend won't yield the full five degrees, but it will be close, in particular with long titanium (low modulus) legs.

View attachment 65741

I knew those Rocket guys were cheating

Oh dang! Dan has uncovered our nefarious plan to use expensive gear legs that behave well. Drats!!