11th may 2024
my plan is to reuse the same burner that i am now going to design for the turbocharger based engine for the axial version too. so much of the design would be for the axial engine. i would also have to design adapters to connect the turbocharger flows to the burner.
i thought that today would be just getting the dimensions of the burner, but i could not be more wrong. seems like, to get the burner right, i need the diffuser passage right, for that, i need the diffuser of the centrifugal compressor passage right. you wiould have noticed that i am writing in present tense right now. that is because the contents of the previous two chapters have been ready even before i stared public logging.
let me start by getting more data at the compressor rotor exit for HX35 for both idle and mil states.
> U: bladespeed
> C: absolute speed
> A: area
| idle | mil | |
|---|---|---|
| rpm | 50000 | 105000 |
| ṁ [kg/s] | 0.204 | 0.429 |
| A exit [m2] | 0.001 | 0.001 |
| Uexit [m/s] | 204.2 | 428.8 |
| Csw_exit [m/s] | 135.1 | 336.1 |
| Cr_exit [m/s] | 119.7 | 160.6 |
| Texit [K] | 320 | 383 |
| Pexit [kPa] | 114.1 | 213.8 |
initially i was worried about shocks because the blade exit speed was way above the speed of sound, but however, since the relative speed of the fluid with respect to the blades are well below mach 1, it will not be a problem.
i will have vanes only after the air has been turned to axial direction. i will use the vaneless space to decelerate the swirl to managable values, i will also make slight changes in the tract passage depth before i could pass them through vanes. i am assuming that when the flow turns, the radial velocity becomes numerically equal to the axial velocity and the swirl becomes unchanged. my goal is to have an axial speed of the order 100 m/s and swirl in the same range. i have two unknowns Rinner and Router.
i have taken my liberty to look at the vaned diffuser characteristic that would be in place when the axial version is made. i have made successfully discovered another set of worries and doubts. i will be mentioning them shortly. here are some numbers i came up with for the vaned diffuser. first i found out the minimum inlet radius required such that the swirl is small enough for my vanes to turn the flow. i set a maximum flow turning angle of around 64° and calculated the inner radius. after calculating the inner radius, i went on to calculate the outer radius as i know the annulus area requirement for the flow to decelerate to the target value. moreover, i do not want any whirl in my combustor. here are my geometries from calculation with the design point being the mil operating point.
> Dinner: 131.1 mm
> Douter: 141.4 mm
> ΔR: 5.148 mm
the vanes will have an inlet angle of 63.46 and 61.69 at hub and tip respectively. however, at lower rpms, for instance at the idle state, the flow has low swirl component and the required angles are 28.33 and 26.5 degrees respectively. i am proceeding that the off design operations will be handled by me in the future. here i list out the burner inlet states at various operating points.
| idle | mil | |
|---|---|---|
| P [kPa] | 131.5 | 365.6 |
| Ptotal [kPa] | 135.4 | 380.0 |
| T [K] | 333.2 | 446.7 |
| Ttotal [K] | 336.0 | 451.7 |
| u [m/s] | 74.6 | 100.0 |
so in essense, our diffuser should handle air entering at the specified annulus at states all the way from idle to mil.
12th may 2024
so many things of yesterdays work were absolute bullshit. i can specify the burner inlet at any resonable dimensions and have the states adjusted accordingly. worrying about the compressor diffuer is another days job. suppose i want the axial flow to have a velocity of 100 m/s at mil, i can also have some swirl at the burner difffuser inlet and have vanes to straight out the flow. there is a slight complication however, in the axial engine, some of the heads would be present as whirl velocity, and diffusion of the whirl component must take place in the burner diffuser using vanes.
i did some numbers for 4 possible states,
> 1. no whirl at mil, 100 [m/s]
here, i will get a static pressure of 367.549 [kPa], static temperature of 446.721 [K], density of 2.867 [kg/m³] with an area requirment 0.00149 [m2]
> 2. 60° whirl at mil, 100 [m/s]in this condition, i will get a static pressure of 326.408 [kPa], static temperature of 431.830 [K], density of 2.634 [kg/m³], with an area requirement of 0.00162 [m2]. since i am desiging for this case as design point. i will freeze this area requirment and find the state for idle operation. there isn't much difference from this area requirement to the one where whirl has been fully converted to pressure head (case 1), it will work for the turbo charger based engine as well.
> 3. no whirl at idle, frozen design from case 2forcing the flow at idle to pass through the above design will yield, a static pressure of 128.608 [kPa], static temperature of 330.950 [K] with a density of 1.354 [kg/m³]. the flow will be at around 100.911 [m/s].
> 4. 60° whirl at idle, frozen design from case 2this would mean a static pressure of 106.698 [kPa], static temperature of 313.761 [K], with desinty 1.185 [kg/m³] at speed 105.846 [m/s]. not that this is only the axial speed. you need to account for 60° of whirl.
using the frozen area, i can freeze reasonable radii so that the gap is not too small. i choose an outer radius of 50.000 [mm] and an inner radius of 44.519 [mm] giving me a reasonable flow tract of width 5.481 [mm]. i hope its enough not too lossy.
live long and prosper,
soorkie