Research Papers: Gas Turbines: Structures and Dynamics

Scaling Concept for Axial Turbine Stages With Loosely Assembled Friction Bolts: The Linear Dynamic Assessment

[+] Author and Article Information
J. Szwedowicz

 Thermal Machinery Laboratory, ABB Turbo Systems Ltd., CH-5401 Baden, Switzerlandjaroslaw.szwedowicz@ch.abb.com

Th. Secall-Wimmel

 Thermal Machinery Laboratory, ABB Turbo Systems Ltd., CH-5401 Baden, Switzerland

P. Dünck-Kerst

 Power Generation/Industrial Applications, Siemens AG, D-47053 Duisburg, Germany

A. Sonnenschein

 Power Generation/Industrial Applications, Siemens AG, D-47053 Duisburg, Germanypeter.duenck-kerst@siemens.com

D. Regnery

 Power Generation/Service Steam Turbines, Siemens AG, D-45473 Muelheim/Ruhr, Germanydavid.regnery@siemens.com

M. Westfahl

 Power Generation, Siemens AG, D-10553 Berlin, Germanymartin.westfahl@siemens.com

All computations were done on IBM computer with two CPUs (EM64) with 8Gbyte RAM with LINUX SLES 9.0 system.

The MECHANICA code cannot be applied for nonlinear problems with large frictional sliding on contacts like the analyzed SK blades.

Degrees of freedom of the nodes on the bolt and airfoil contact area (Fig. 1) are tied rigidly in the normal and in two tangential contact directions.

Only in the normal direction of the contact, degrees of freedom of the nodes on the bolt and airfoil contact area (Fig. 1) are tied rigidly.

Because of spatial localization of the vibration energy, certain airfoils, which are disordered by manufacturing and assembling tolerances as well as by divergences in material properties, might experience substantially larger oscillations than the numerical response amplitudes of the tuned bladed disk.

J. Eng. Gas Turbines Power 130(3), 032504 (Apr 03, 2008) (12 pages) doi:10.1115/1.2838995 History: Received July 03, 2007; Revised November 09, 2007; Published April 03, 2008

In the early 1980s, Siemens developed a last stage fast rotating condensation blading (SK) blade with strongly twisted and tapered profiles for industrial condensing steam turbines, which operate with variable speed under high steam mass flow and excessive condensing pressures. To suppress alternating stresses of the lowest blade resonances, conical friction bolts are loosely mounted at the upper parts of adjacent airfoils. Also, these bolts couple the rotating blades, since steam excitation is lower than the friction threshold force on the bolt contacts. These coupling and damping capabilities were proven experimentally for the smallest SK blade at the test rig of the real turbine. By considering the similar mechanical and aerodynamic characteristics based on the tested smallest airfoil, the entire SK-blade family has been scaled up for reliable utilization in more than 500 industrial turbines operating for diverse ranges of power and speed. A recent trend to very large compression units, like gas to liquids, acid terephtalic, or methanol plants, imposes a need for further enlargement of the SK-blade family and its friction bolt, whose mechanical properties have been proven experimentally for the smallest airfoil. In this paper, the mechanical capabilities of the smallest and large SK blades coupled by the bolts are verified by using the finite element (FE) method. The static analyses with friction sliding on airfoil interfaces and the linear dynamic behavior of tuned disk assemblies are considered. The FE mesh quality and the proper boundary conditions at the radial fork root are accomplished by getting good agreements between the computed and measured resonance frequencies of the large freestanding blade at standstill. The validated mesh refinement and root boundary conditions are used further in all numerical FE analyses. For the large SK-disk assembly under spin-pit conditions, the obtained FE results are in very good agreement with the experimental Campbell diagrams, which are measured with the three gauges that also identify the stick-slip and stuck bolt’s contact conditions. Concerning the gauge outputs and the FE steady-state blade resonances computed for the analytically determined air jet excitation, the experimental spin-pit results demonstrate that the bolts are mainly in stuck contact conditions. Only in very narrow frequency ranges around resonance peaks, microslips on the bolts occur due to the resonance amplification of blade vibrations. This is proven indirectly by comparison of the overall damping values evaluated from the blade resonances at standstill and in the spin pit. The described linear dynamic concept assesses properly static stresses and free vibrations of the scaled disk assembly with friction bolts. For the steam excitation, which generates dynamic contact reactions bigger than the friction threshold forces, the realistic blade responses need to be obtained from the blade simulation with friction (Szwedowicz, J., Secall-Wimmel, T., and Duenck-Kerst, P., 2007, “Damping Performance of Axial Turbine Stages With Loosely Assembled Friction Bolts; the Non-Linear Dynamic Assessment; Part II  ,” Proceedings of ASME Turbo Expo 2007, Montreal, Canada, May 14–17, ASME Paper No. GT2007-27506).

Copyright © 2008 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Figure 1

(a) The last LP stage of industrial Siemens steam SK turbine with (b) friction bolts, where the angle φ=2π∕N illustrates the geometrical periodicity of the disk sector for N number of blades in the row; (c) the considered scaled-up SK blades

Grahic Jump Location
Figure 2

Cyclic sector model representing one SK blade, the rotor part, and one bolt divided into two symmetrical halves

Grahic Jump Location
Figure 3

(a) Siemens’ hammer setup with the airfoil measured at (b) 100 points for the mode shape identification of the large SK blade with (c) “long” and (d) “short” roots

Grahic Jump Location
Figure 4

(a) FE model of (b) Siemens’ setup for hydraulic clamping of the airfoil root on its circumferential sides

Grahic Jump Location
Figure 5

(a) Siemens’ setup for hydraulic pressing of two guide bolts to the second and third airfoil fingers of the radial root; (b) the FE normal contact stresses, where blue color refers to noncontact state

Grahic Jump Location
Figure 6

Relative errors of the FE blade frequencies computed with the contact root model illustrated in Fig. 5 in relation to the measurements at the hydraulic setup (Fig. 5) and at standstill (Fig. 7) before spin-pit tests

Grahic Jump Location
Figure 7

The experimental resonance frequencies of eight large SK blades assembled to the rotor without bolts at standstill before the spin-pit test (T=20°C)

Grahic Jump Location
Figure 8

Comparison of the centrifugal stresses between the smallest (a) and large (b) blades coupled by the friction bolts at their nominal speeds and T=20°C

Grahic Jump Location
Figure 9

Computed final clearances at the ends of the bolts for (a) their maximum and (b) nominal dimensions of the large and small blades at their nominal speeds

Grahic Jump Location
Figure 10

FE normal contact stresses on the bolt of the large blade at (a) 100% and (b) 25% of its nominal speed Ωn

Grahic Jump Location
Figure 11

Coupling bolt effect demonstrated by a frequency increase of the large SK disk assembly at 50% and 100% of the nominal speed n≡Ωn, where ωr,1, ωr,2, and ωr,3 are the resonances at Ω100%

Grahic Jump Location
Figure 12

The numerical Campbell diagrams of the freestanding (fi is the ith eigenfamily) and coupled large SK blade in the spin pit, where fi,n means the eigenfamily i vibrating with the nodal diameter n, and k={1,2,…,∞} is engine order

Grahic Jump Location
Figure 13

Validation of the frequency scaling factors determined from the FE spin-pit Campbell diagrams of the smallest and large SK blades (Fig. 1), where the design scaling factor υ0 equals 4

Grahic Jump Location
Figure 14

Three lowest mode shapes of the blade coupled by (a) sticking and (b) slipping bolts, where SG#2 indicates the position of Gauge 2. Parameter E33MAX means the relative strain magnitude in the radial direction determined by {ε°}i,n=[{ε°i,n}c+j{ε°i,n}s]1∕2

Grahic Jump Location
Figure 15

(a) Position of Strain Gauge 1 (SG#1) instrumented on four adjacent blades for measuring the disk vibrations up to the 15th engine order (Fig. 1) and (b) Location of three gauges on four adjacent blades

Grahic Jump Location
Figure 16

The measured Campbell diagram of the four lowest eigenfamilies (ωC,i=1,…,ωC,i=4) of the large SK blades excited by air jet in vacuum at 20°C, where ΔΩC is the speed region while the bolts begin to couple the blades

Grahic Jump Location
Figure 17

Order tracking of harmonic excitations of (a) k=7 for resonance i,n=1.7, (b) k=6 for resonance i,n=1.6 (Fig. 1) with the measured double resonances ω′ and ω″ due to mistuning effects, where Δω=ω″−ω′

Grahic Jump Location
Figure 18

The comparison of the computed and measured Campbell diagrams of the large SK blades, where the FE eigenfrequencies fi,n refer to eigenfamily number i={1,2,3} and nodal diameter numbers n={0,1,2,3,6,9,12,16,19}

Grahic Jump Location
Figure 19

The numerical and experimental Campbell diagrams for resonances of the first family excited by engine orders k of 6 and 7, where γ1,6 is the degree of blade coupling for disk mode i,n=1.6

Grahic Jump Location
Figure 20

The numerical and experimental Campbell diagrams for resonances of the second and third eigenfamilies, where γ2,11 and γ3,12 are coupling degrees of disk modes i,n of 2.11 and 3.12

Grahic Jump Location
Figure 21

(a) SK turbine train in the spin-pit chamber and (b) the pipe arrangement for the airfoil excitation, where κ is dimensionless polytropic constant and Tt, pt, ρt, and vt are the temperature, atmospheric pressure, air density, and flow velocity outside the bunker

Grahic Jump Location
Figure 22

(a) The air jet excitation force f(α) and (b) blade mode shapes ϕ1.6, ϕ2.11, ϕ3.12, where α1 and α2 are angles at the entry and exit of the blade contours by the air gust at time t1 and t2 during one period T=2π∕Ω of the turbine rotation

Grahic Jump Location
Figure 23

Excitation amplitudes Fk and phase delays κk of engine orders k of the air jet excitation f

Grahic Jump Location
Figure 24

A waterfall diagram of the spin-pit measurement of the large SK blade at strain gauge SG#1

Grahic Jump Location
Figure 25

The computed (the overall damping ξ=0.14%) and measured strain responses at SG#1 gauges on eight blades in the spin pin (Test Runs 08, 14, 20, and 41) for resonances (a) i,n=1.6 excited by k=6, (b) i,n=2.11 excited by k=11, and (c) i,n=3.12 excited by k=12




Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In