Roee Kalinsky's RV-7A Project

2009.08.26: (0.0) This documents what is essentially a complete redesign that I did of the brake fluid reservoir assembly on the RV-7A.

THE PROBLEMS

The brake fluid reservoir supplied in the kit is a TIG-welded aluminum cylinder with a pair of mounting flanges and an aft-facing bung, and is mounted near the top starboard corner on the forward side of the firewall.  The installation is shown on DWG ???.  The general arrangement is ok, but there are many minor problems with the details:

1. The mounting flanges are welded to the cylinder wall and essentially form long and relatively weak cantilevers.  It is easy to imagine that with normal flight loads and vibration, these welds may eventually fail, and indeed some RV owners have reported this occurring.

2. The top of the firewall angles forward and interferes with the top of the reservoir, therefore requiring that washers will be added behind the reservoir mounting flanges to shim it further forward and away from the firewall.  This further increases the potential for vibrational loads on the flanges.

3. The bung at the bottom of the reservoir is an aft-facing tube that has female NPT threads to accept a fitting.  The problem with this arrangement is that the tube has to penetrate the firewall,  but no good means for mechanically securing or sealing the penetration are provided.  In essence, one has to drill a clearance hole through the firewall sufficient for a slip fit for this tube, and then goop some sort of sealant (proseal, hi-temp RTV, etc.) all around it.  This is not robust as a mechanical bond nor as a gas seal in the long term, especially given that normal vibration will tend to cause relative motion between the reservoir bung and the firewall in this area.

4. The reservoir has only a single bung, whereas it ultimately needs to feed both the left and right brake systems.  Naturally, a "T" fitting can be used.  But this makes the arrangement prone to working itself loose over time, since normal motion of the rudder/brake pedals will transmit tortional forces back to the "T" fitting via the hoses.

POSSIBLE SOLUTIONS

Not satisfied with Van's reservoir installation, I decided to redesign it to eliminate these deficiencies.  I looked at several options:

1. Eliminate the firewall-mounted reservoir altogether and use two smaller reservoirs mounted directly to the co-pilot's master cylinders.  This arrangement has been used in some production aircraft, as well as some RV's, and works reasonably well.  The biggest downside I see is the potential for brake fluid spillage in the cockpit, both during maintenance service (i.e. bleeding the brakes, checking or topping fluid levels) and during high tubrulence or other momentary negative-G conditions.  [Note that in all the arrangements discussed, the top of the reservoir is vented through a filter that should prevent spillage under only very brief negative-G conditions, such as may be incountered in moderate turbulence.]

2. Design and build my own "ideal" firewall-mounted reservoir.  I ruled this out only because I don't currently have access to a machine shop.

3. Buy a different (i.e. 3rd party) firewall-mounted reservoir and design an appropriate installation for it.  Ultimately this is the option I chose.

MY SOLUTION

I looked at several off-the-shelf reservoirs from several manufacturers, and selected one from Grove (p/n 067-054) as the best fit for my needs, or at least the best starting point.  It is a machined aluminum cylinder (6061 alloy, black anodized finish) with a bottom-facing female NPT outlet.  It comes with a mounting bracket that connects to the reservoir itself using two 10-32 screws, which is nice and sturdy, but that also gives me the freedom to make a different mounting bracket if necessary (and it was).

In designing the installation, I wanted to keep the same general location on the firewall as Van's original design, but eliminate the problems listed above.  In essence, I opted to make the "T" on the forward side of the firewall, and then have two separate penetrations through the firewall using AN bulkhead fittings.  Also, I opted for everything (the reservoir and the fittings) to be secured to a single rigid support bracket that could then be attached to the firewall as a unit, thereby eliminating any relative motion due to vibration of the firewall.  I played around with a whole bunch of different ideas and configuations, and ultimately landed on the following.  A picture is worth a thousand words, so here's several thousand words worth.

Van's reservoir (left) and mine during initial fitting (right):



Installed on the firewall:






And with the engine mount installed.  Hard to judge from the photos, but the installation (reservoir, tubing, etc.) has better than 1/4" clearance all around from the engine mount as well as the firewall, ensuring that nothing will ever come in contact even with engine vibration.





PARTS

The mounting bracket I fabricated from 6061-T6 aluminum, 0.125" thick 2"x2" angle stock.

The reservoir itself, as mentioned above, is Grove p/n 067-054.

Instead of the mounting screws and split washers supplied by Grove, I intend to use AN3H-3A drilled head bolts and AN960-10 washers, and safety wire.

The fluid fittings are all standard AN hardware (with one exception**):
One AN822-4D elbow, 90°, 1/8" NPT to 1/4" flared tube.
One AN804-4D bulkhead tee, 1/4" flared tube with bulkhead on the run.
One AN833-4D bulkhead elbow, 90°, 1/4" flared tube.
One 818-4-4D** straight female-to-female swivel coupling, 1/4" flared (**AN-compatible but not a standard AN part, purchased from Bonaco).
All the usual AS4824 conical seals, AN819-4D sleeves, AN818-4D and AN924-4D nuts.

The tubing I fabricated myself (5052-0 aluminum 1/4" aluminum tubing) and has the following dimensions: 1.19" throat (i.e. center to center bend diameter) and 0.9" difference in length of the straight sections.  The bend radius is simply that of my Imperial tubing bender, and the length of the straight sections would ideally be as short as possible (given the 0.9" difference), but were dictated by limitations imposed by the design of the tubing bender and flaring tool that I have.  So in a sense, much of the rest of the installation geometry was designed around that, but all worked out well.

THANKS

Thanks to the folks at Grove for putting out a very high quality product, well designed and well made, as I've come to expect from them.  And thanks to Brett Jarvis at Bonaco for all his good suggestions and for sending me parts very quickly when I was ready to do it (we spoke in the late afternoon, UPS was at my door the next morning).  Both Grove and Bonaco will certainly be getting more of my business in the future.