,

 The trim number is calculated thusly Trim=(Inducer diameter squared divided by the Exducer diameter squared) x 100

This is where determining the trim of the wheel is critical to the design.  A smaller trim wheel means the inducer diameter is smaller.  Another way to look at this is the leading edge of the blade is shorter.  You can liken this to a cantilevered beam.  Because the blade/beam is shorter, it does not have to be as strong.  Going to a bigger inducer diameter means a longer leading edge, or think of it as a longer cantilevered beam.  Because it is longer, there is more leverage forced on it which means the blade has to be stronger and thicker like a longer cantilevered beam has to be beefier.  Basically, a bigger trim compressor wheel has to be made stronger than a smaller trim wheel.  One way this is done is by making the blades thicker.  Also, in the area where the blade meets the hub of the wheel, the fillet radius can be optimized to improve strength.

 On a bigger trim wheel, the inducer diameter is larger, the blade tip here is further from the hub giving it a longer leverage arm to create more stress at the root of the blade, thus a larger trim wheel must have stronger and thicker blades.

To address LCF and HCF during the design process, an FEA analysis is performed to determine the harmonics of the blades.  Each order of harmonics creates stresses in different regions of the blade by causing the blade to bend in different ways.  In the regions of high stress, the blade geometry and thickness has to be optimized to prevent failure.  How can you tell when a design is not strong enough?  The blades literally break off!  The starting point is generally the tip of the inducer blade as it is a high stress region and it will break off the blade.  This may look like FOD (foreign object damage), but it’s really just the blades being too weak.  Not only is this not good for aerodynamic performance, but engines typically don’t like ingesting bits of aluminum either.

 Turbos spin so fast that even 7th order harmonics have to be considered off of the blade's natural frequency.  Each order of harmonic will stress the blade differently and every change to the blade will affect the harmonics.  That's why computer modeling can greatly help with blade stress analysis.

After performance and strength, the last hurdle to jump is actually making the wheel.  If the wheel is designed as a casting, the shape of the blades has to be pullable.  A blade shape is no good if after you pour the metal to make the wheel, you can’t get it out of the tool!  Going to a fully machined wheel on a 5-axis machine eliminates this limitation on blade geometry.  Perhaps even more importantly, fully machining a wheel means it can be made from a forged piece of aluminum for increased strength.

Not all billet wheels are created equally.  Unlike wheels from some other manufacturers, the Garrett GTX wheels are machined from a forged chunk of aluminum rather than rolled round bar stock.  Without the forging process, the metal is vulnerable to inconsistencies and weaknesses. Forging improves the grain making it finer and more homogenous.  The mechanical working of the forging also improves the tensile strength of the material.  Very importantly, Garrett's forged wheel blanks are close to the net shape of the wheel, mirroring the outline shape of a wheel.  By doing this, the grain structure gets aligned into the basic shape of the compressor wheel greatly improving its strength.  If you cut in half a GTX wheel and etch it to see the grain structure, you can see that the grain structure actually follows the shape of the wheel.

 Garret forged blanks are close to net shape so the grain flow is all around the actual shape of the wheel.  A lot of other billet wheels are cut from bar stock, thus the grain flow is not quite as good.