Sunteți pe pagina 1din 8

MECH3404

Assignment 1
Vibrations and Signal Processing

Arti Pillai, Ishaan Bahree, Antriksh Srivastava, Victoria Sita 5/2/2012

Aim: Our focus was to minimise the maximum response of a clamped/free bar through the addition of a torsional vibration absorber depicted below. It was necessary to optimise the length of the absorber, L2, and the value of the spring constant k in order to achieve the minimum maxima of the response of the free end of the bar concerned.

Background Information: In order to design the torsional absorber, in depth research of vibrational response was conducted aiding and supporting our calculations. We realised it was necessary to divide the system into two major systems, the undamped clamped/free bar system and the torsional vibration absorber system. It was understood that by adding the effects of both systems, the total receptance could be found, the minimum maxima being the ideal solution. Receptance can be defined as the response of a system in a particular direction over an excitation force that has been applied to the system. For different systems, whether clamped or free bars, the receptance will change respectively. System 1. Clamped/free bar:

The receptance for clamped/free bars can be calculated using the equation below.

System 2. Free/free bar:

The receptance for free/free bars can be calculated using the equation below.

For the case concerned, our free/free bar contains two parts, the rubber insert that behaves as a hysteretic damping element with a specified characteristic and the actual free/free bar that behaves as a mass. The total receptance of the second system will hence be the sum of the receptances of both the rubber insert and the free/free bar. The diagram below shows how to find the overall receptance of the system:

The overall receptance is the reciprocal of the sum of the inverse receptances of both System 1 and System 2 following the equation below.

Where;

is the total receptance is the receptance of System 1 is the receptance of System 2

Assumptions and Values: Our calculations are based on the assumptions that the the torsional vibration absorber acts as a free/free bar and a spring mass with a rubber insert behaving as a hysteretic damping element characterised by k(1+i). The values given are as follows. L1 = 0.3m D1= 0.15m D2 = 0.15m = 0.2 = 7800 kg/ G = 8x N/ Calculations: For n=1 we get 16768.6 rad/sec, and for n=2 we get 50305.74 rad/sec Using this equation, we can find the first two natural frequencies by substituting 1 and 2 into n. For the L we replace it with L1 since it is the member we are analyzing here. The other L2 is simply the length of the mass of the spring and hysteretic damper. For = 16768.6 , for = 50305.74 The equation above shows how to find the wavelength. The density and the Shear modulus is constant and only the natural frequency has two values which is determined in the previous equation.

The beta here is the receptance of the first bar which is clamped free. The lambda is the only variable which takes two values as found before.

The gamma here is the combined receptance of the hysteretic damper and spring. The initial part of the equation before the minus sign represents the spring and the damper. The part after represents the receptance of a free/free bar.

Given, Therefore;

This equation is the inverse sum of both beta and gamma. Alpha here is the overall receptance. The sequences can be added in series.

| |

This is the real component of

| |

Above is the imaginary component of After finding the real and imaginary parts of

we can combine them by:

The process above has been completed on MatLab since it is very complicated to do by hand.

MatLab:
Matlab Script: Since the equations we derived were complicated and involved complex numbers, we used two matlab scripts to solve the problem. The first script, function [r] = theFunction(k, L2), sets the equation to solve for total receptance by using the values given on the assignment and the unique parameters given to our group. The equations we developed were then input into the script to solve for the total receptence at each frequency value between 0 and 70000 rad/s. The next step was to minimise the maximum response over the frequency range and find the corresponding values of k and L2. This was done writing the script, function loopThrough() Which uses a while loop. A while loop allows the user to set parameters to which matlab will continue to solve a function for, until the optimal solution is found. We set the loop to run for k values between 0 and 20000000 while L2 remains less than 0.3m (unique parameter and condition stated in problem). Running the script solves for k and L2 until the response is minimised. The fprintf('Max= %e. k = %f, L2 = %f.\n', temp, k, L2) command prints out all the solutions of k and L2. Since the script runs for so many values of k we have not included all the print outs below. Once the optimal values of k and L2 are found the script outputs these values.

Lastly, we used the plot function to create a subplot showing the undamped response of the bar and then the damped response of the bar. The undamped response plot shows the behaviour of the bar before it was damped and the damped response plot shows the behaviour of the bar after damping. function [r] = theFunction(k, L2) %Variable declarations p = 7800; (kg/m^3) L1 = 0.3; G = 8*10^10; J1 = 4.97*10^(-5); J2 = 4.97*10^(-5); w = 0:1:70000; %Lambda function l = (w) * sqrt(p / G); F1 = sin(l * L1) ./ (G * J1 * l .* cos(l * L1)); F2 = -cos(l * L2) ./ (G * J2 * l .* sin(l * L2)); Receptance F3 = 1/(k*(1+0.2i)); Receptance F4 = F3 + F2; temp = 1 ./ ((1 ./ F4) + (1 ./ F1)); Alpha r = abs(temp); alpha %Minimising Maximum Response function loopThrough() k = 0; L2 = 0.01; w = 0:1:70000; maxResponse = 1; optimalL2 = 0; optimalk = 0; %Beta Receptance %Free/Free bar %Spring/Damper %Gamma Receptance %Total Receptance, %Absolute value of

%Density of the material %Length of L1 (m) %Shear Modulus (N/m^2) %Frequency range

%Length of L2 (m) %Frequency Range

while k < 20000000 L2 = 0; while L2 < 0.3 temp = max(theFunction(k, L2)); if (temp < maxResponse) maxResponse = temp; optimalL2 = L2; optimalk = k; fprintf('Max= %e. k = %f, L2 = %f.\n', temp, k, L2); end L2 = L2 + 0.01; end k = k + 10000; end %Plotting the Response of the bar %Variable declarations.

p = 7800; (kg/m^3) L1 = 0.3; G = 8*10^10; J1 = 4.97*10^(-5); J2 = 4.97*10^(-5); w = 0:1:70000; %Lambda function. l = (w) * sqrt(p / G);

%Density of the material %Length of L1 (m) %Shear Modulus (N/m^2) %Frequency range

F1 = sin(l * L1) ./ (G * J1 * l .* cos(l * L1)); F2 = -cos(l * L2) ./ (G * J2 * l .* sin(l * L2)); receptance. F3 = 1/(k*(1+0.2i)); Receptance. F4 = F3 + F2; temp = 1 ./ ((1 ./ F4) + (1 ./ F1)); Alpha. r = abs(temp); alpha. subplot(2,1,1) Frequency plot(w,abs(F1)) grid on xlabel ('frequency w') ylabel ('Beta') title ('Torsional Response of Bar - Undamped')

%Beta Receptance. %Free/Free bar %Spring/Damper %Gamma Receptance.

%Total Receptance, %Absolute value of

%Plots Beta against %abs of F1(beta)

subplot(2,1,2) %Plots Response against Frequency plot(w,r) grid on xlabel ('frequency w') ylabel ('Total Response, Alpha') title ('Torsional Response of Bar - Damped') fprintf('Optimal k = %f, Optimal L2 = %f.\n', optimalk, optimalL2);

Running the matlab script with the given inputs, we got the following results: >> mech3404assignmentNew11 Max= 1.223947e-003. k = 0.000000, L2 = 0.010000. Max= 4.776680e-004. k = 10000.000000, L2 = 0.010000.
Max= 2.418422e-004. k = 20000.000000, L2 = 0.010000. . . .

Max= 4.498412e-007. k = 9980000.000000, L2 = 0.220000. Max= 4.494265e-007. k = 9990000.000000, L2 = 0.220000. Max= 4.490126e-007. k = 10000000.000000, L2 = 0.220000. Max= 4.485997e-007. k = 10010000.000000, L2 = 0.220000.

Optimal k = 10010000.000000, Optimal L2 = 0.220000.


1.5 1 x 10
-3

Torsional Response of Bar - Undamped

Beta
0.5 0 0 x 10
-6

3 4 frequency w

6 x 10

7
4

Torsional Response of Bar - Damped

Total Response, Alpha

0.5

3 4 frequency w

6 x 10

7
4

Conclusion: As seen from the resulting plot above, the maximum value of the peaks have been drastically reduced suggesting that the k and L2 values indeed induce optimum damping.

References: B.J Stone, J.Pan, Vibration Causes and Cures, Chapter 9, Avialable at [http://school.mech.uwa.edu.au/~bjs/eVib/Receptances.pdf, Accessed 27th April 2012] B.J Stone, J.Pan, Vibration Causes and Cures, Chapter 7, Avialable at [http://school.mech.uwa.edu.au/~bjs/eVib/Continuous.pdf]

S-ar putea să vă placă și