Monday 26 February 2018

Muffler Cluster @%#^

This blog covers what called a Cluster !#% begin supplied as a muffler system.

Overview
It was decided to save some muffler weight by selecting a titanium muffler manufactured by Akrapovic. Though expensive it carved two [2] kg from the total weigh and was very compact in design so on its arrival it was found to be of a first class construction and fitted like a glove and confidence was high.

Later examination revealed a rather small perforated inlet to the exhaust side of the muffler but a internet search yielded no bad reports but it was put on the to look at list. A recent Facebook Post created a reply from Thomas at Edge Performance indicated that this was indeed a problem so it moved up my to do list. 

Examination of the inlets and exhaust with a torch showed a perforated inlet of at least four times the length of the outlet with the areas begin larger than the inlet/outlet pipes based on a 50% open area. However Thomas comment concerned me so the decision was made to open up the outlet for possible modification and this is were the reason for the failure of the design became slowly apparent.


Exhaust outlet - No I am not kidding
that's all there is

With the access the builder decided to grab the torch and look up at the inlet and that is when another issue revealed its ugly head, that 100 mm perforated tube had only 50 mm of exposed area as 25 mm was mounted in a flared slip tube to reduce thermal stress.

At this point the logic of the system escaped me totally - OK I am not that bright but this just left me lost for why..!


A - the slight shadow id the end of the flared mount
B - Is the 50 mm of open tube
C - The end of the inlet tube

A lot of time was spent on the internet with no comparable concept found until a clip on why a turbo muffler can flow more gas that the classic hot dog straight through. Classic turbo mufflers form a looped path with the inlets and outlets offset from each other - So why are they more efficient that a straight line of the hot dog?

The answer lied in the fact that the perforated material of the legs inside the muffler had a ever increasing boundary layer even though some gas escaped via the perforated leg until it exited to the next leg creating an increase in gas velocity.






So in a previous life I had been involved with piping systems and it just hit me - this fool has effectively coupled up both headers [inlet 1 /2 image below ]. Let me explain - fluids all have weight therefore inertia so the concept of stopping a discharge in 50 mm while turning 90 degrees before the next charge arrives is not plausible in any world.

Looking at one cycle with unequal header lengths feeding the muffler, the first charge [short header] enters the muffler with a balance of the gas accelerating past the perforated tube up the long header until the inevitable collision occurs. Now the cycle is repeated in the opposite direction with evacuation more of a result of the pressure build up in the headers.

I tend to never criticise other peoples designs but this is a total cluster stuff.

If this description is true there is only one solution and that's a 1970 hot rod muffler gutting. The edge is bead welded so I will try to grind off the bead and then the ends should be removable allowing the guts to fall out. Stubs will be welded to each inlet and the inlet cut at 30 degrees to create a smoother acceleration as it enters the exhaust wit the jury out on the inlets.

As this aircraft will fly as experimental in Australia it is in Phase 1 testing all its life so no noise restriction applies other than a social one. 


Proposed design - bit rough but a first cut

Comment
None - I am at a loss unless some one can explain this to me in simple terms

Sunday 4 February 2018

Cowls - Part 3

This blog covers the fitting the lower cowls cheeks

Overview
With the top cowls ready to be installed the lower cowls now had to be fitted in a similar  manner to the top.

Installation
The starboard side was fitted back into the flange jig and clamped to the side rail of the upper cowl attachment. A digital protractor was zeroed on the fuselage side, placed on the flange of the cowl and leveled.








With the cowl located a light was placed on the inside and the edge of the fuselage marked on the outside face of the cowl. 

With the edge marked a offset line of 2 mm was added and the waste trimmed using a Dremel motor tool fitted with a thin cutting disc.

The next task was to fit a fastener mounting to for the lower cowls. With the experience  from the upper cowls a aluminium angle was machined using the proven pattern.

With the sides trimmed a series of slots was filed to expose the fuselage edge as outlined in Part 2 of the cowls.

With the edge adjusted to fit the first mounting point was drilled 45mm [1.75''] down from the last mount for the top cowl. Six [6] more were added pitched at 90mm [3.5''] as per the suppliers recommendations.


First lower cowl attachment with
lower flange attachment fitted

Next task was to install the lower top cowl attachment that will tie the two assemblies into one unit. 

The Tucano cowls have a flange on the side supported by a integrated angle, to use the Skybolt fasteners requires an aluminium support to achieve the required strength, so the flange above the angle was removed with a Dremel Motor tool.


Cowl after trimming shown in insert above
Aluminium mount shown in o/a imiage

Next a new aluminum angle was machined by notching with the 10mm ball mill pitched at 45 mm along the full length of the lower flange. This was fixed using 3mm pop rivets fitted with 3mm brass washer on the underside to prevent the rivets pulling through.


Aluminium angle trial fit before riveting

The same techniques were applied to the port side and when the level across the two cheeks was checked it is was less than 0.5 degrees.

Comment
When a little smoother for the firewall angles while the cowl top edge was as we say "a drop kick". The way the aluminium flanges met were the hand of God as no one was more surprised than me..!

Next job is to glue the lower cowls and to trim and install the top cowl