Effect Step and Gaps has on flow over a surface

[SafeTech]

Hi

I would like to evaluate the impact steps and gaps has on the
aerodynamic performance. At our company we design and build exhaust nozzles for large aircrafts. As with any sheet metal part sometimes the weld height or steps and gaps, surface waviness exceeds allowable limits. The impact structurally is small however, do anybody have or know where I can find out on the aerodynamic impact. Does this require CFD analysis. If so, which package you would recomend running on the PC.

Thanks

Aston

 

[insideman]

A fantastic source for this sort of information is "Fluid Dynamic Drag" by Sighard Horner. I got my copy from Powells books for $80, free shipping.

 

[WKTaylor]

SafeTech

Be very careful about steps INTO exhaust airflow... they can protrude thru the boundary-layer and become "hot-spots". Also, they and can alter back-pressure slightly... not to mention destabilize the boundary-layer and re-vector the flow center-line. The effects can be small or sizeable depending on the exhaust diameter-to-step and step-to boundary-layer ratios. Total thrust can be affected; and the engine may be more susceptible to thru-flow problems [compressor stall, surges, etc].

Steps AWAY from the boundary layer can also prove trouble-some, since depressions can contain entrapped vortexes... also altering flows and back-pressures and thermal-distribution... tho less of a problem if small-area.

Contour deviations can induce unwanted stresses on exhaust-ducts at joints. Ducts see radial-pressure loads, "pull-off" loads [tension due to exhaust friction drag], gas-flow errosion, thermal and sonic/vibration loads ... just to name a few of the "biggies". Deviations from contour can affect the durability of the duct... especially along weld-lines with sharp steps due to cracking or creep-distortion. Certain material are more tolerant... others less tolerant.

Exhaust duct contours, joints and fabricating processes need to be VERY carefully controlled... they will see plenty of hasrsh service... and need to start-out as close to BP as possible to survive and perform well with the installed engine.

Regards, Wil Taylor

 

[creative]

I have a similar query with regards to effect of surface waviness on surface flow. This time, my problem pertains to flow into the engine intake.

I read in aviation books that engine intake ducts are painted, possibly with a polyurethane paint coating. However, friends have warned about erosion of this paint due to intake shock/flow, thereby creating surface roughness and affecting boundary layer flow.

I would like to know if CFD can be used as a tool to model possible surface roughness, or the kinds of qualification tests that companies should carry out to prove either (1) that shock (pressure/temp gradients) and high velocity airflow do not affect the paint or (2) engine is tolerant to minor surface waviness (scale of 1mm ht variation vs a 600mm diameter round duct).

THanks

 

[Haf]

Subsonic or supersonic flow over a wavy wall (the wavy wall problem) is a classical fluid dynamics problem that is usually covered in graduate level courses (you may find it in some compressible fluids textbooks; try "Compressible Fluid Dynamics" by Hodge and Koenig). Start by assuming the flow has periodicity. Simply estimate the amplitude (1 mm for your situtation) and wavelength, and write the surface equation as a sin/cos function. If the amplitude is much less than the wavelength, you can use the small disturbance equation, which is a linearized version of a much more complex equation.

For subsonic planar flow, the equation is elliptic and similar to Laplace's equation. It can be solved using the separation of variables technique. For supersonic planar flow, the equation turns into the wave equation, and d'Alembert's solution can be used. If you are dealing with axisymmetric flow, the equations change a bit, and solutions become a bit more complex. For example, for subsonic flow, separation of variables can be used, but then Bessel functions are needed.

Solving the small disturbance equation will give you the perturbation velocity potential, which can be used to calculate things like perturbation velocities and wall pressure coefficients. This data should help your engine folks determine if the engine can handle the perturbances.

If you don't want to deal with solving this problem by hand, these equations could easily be incorporated into a CFD program.

As for whether or not shock wave interactions affect the paint, I don't think I can help you there. The only thing I can think of is to run some experiments in a wind tunnel. The above analysis, however, should help out with determining if there is a significant effect from surface waviness and/or how much surface waviness your engine can tolerate.

Haf

 

[WKTaylor]

creative...

Ah I have lots of experience in Inlets!

The radiused intake-lip is critical... and so is the area just forward of the engine inlet-plane.

Transport/corporate jet-acft typically have short inlet-ducts. The intake-lip is followed by a short section of skin to the engine-inlet plane. This section of skin is usually perforated to bleed-off shocks and pressure surges.

Fighter/stealth acft with long inlet ducts have a similar critical areas... but between the intake-lips and the perforated skin forward of the engine-inlet plane is solid skin-ducting. This duct gradually transitions from the rectangular or "D" shape of the intake-lip cross-section to the engine-inlet-plane cross-section [round].

In supersonic fighter ACFT [F-15, F-16 and F-4] I have never seen inlet duct coating failures that were not explained by (4) factors:
(a) poor surface preparation... leading to loss of coating adhesion.
(b) Loose fasteners... leading to loose joints and fasteners-head... causing paints and sealants to chip and flake.
(c) Engine Stall/Stagnations that were severe... leading to significant duct-strains. The duct skin flexed around the fastener heads causing paint separation/bubbling. NOTE: stall stag limit over-pressures can reach 60PSIA!
(d) FO impact damage to the coating.

Regarding protrudences in the solid-skinned duct section: I have had NUMEROUS scab patches installed that were “sized” for the skin thickness… and had various odd shapes and sizes. Some semi-rectangular patches spanned 1/4 to 1/3 of the inlet “circumference”!!!! The edges were generously tapered all-around... and they were installed “wet” with sealant, which was filleted to the patch edge. "Experimentally" we saw NO degradation in engine performance. We did observe erosion on the (chamfered) leading edges of the patch and the sealant fillets... but usually nothing serious... unless there was poor pre-treatment prior to painting!

Anodize/Alodine, Epoxy-Primer and Polyurethane top-coats prove to be very tough/durable combination… able to withstand the low/high temps of engine inlets!
Regards, Wil Taylor

 

[creative]

Thanks Haf and Wil.

Wil, with regard to the patch-tests which you carried out, may I know the thickest patch that was installed? Also, what is the length of the duct that you tested and what was the max thrust of the engine?

 

[WKTaylor]

Creative..

The areas we patched in the F-15 inlet-ducts were about 28 to 30-inch max cros-section [usually oval or rounded square. Typical patch thicknesses were 0.07--0.08" thick [for fastener head countersink].
Regards, Wil Taylor