1. BUDAPEST UNIVERSITY OF TECHNOLOGY AND ECONOMICS ELECTRONICS TECHNOLOGY DEPARTMENT Final Report 04.15.2009 Rev.: 22.05.2009 Dr. László Jakab, László Milán Molnár, Olivér Krammer FEM simulation of stencil deformation

2. BOSCH - STENCIL_FEM 2/59 WORKPLAN OF THE PROJECT Phase one – simulation - measuring the necessary parameters – done - collecting the missing technical data – done - constructing the FEM model for the PCB – was not necessary according to kick-off meeting Milestone 1 – simulation model of the PCB is ready – pilot stencil model was ready instead of the PCB model - measuring the elastic properties of stencil – done - constructing the FEM model of the stencil – done, extended with the FEM model of a real squeegee - running simulations – done Milestone 2 – Decision about the continuance depending on the results of the simulations – project is continued Deadlines from the start of the project (15.10.2008) End of 2 nd month End of 4 th month Phase two – Designing and carrying out experiments - setting up experiment – testboard is ready - carrying out stencil printing tests – done - measuring printing transfer efficiency – done Milestone 3 – Comparing simulation and experimental results, evaluation the results of the project – done End of 6 th month

3. BOSCH - STENCIL_FEM 3/59 CONTENTS 1.Investigating stencil deformation in case of point loading: Stencil is loaded at the center; different sizes of underside supports were used; pilot FEM model of the stencil was created according to deformation results. 2.FEM model of a real squeegee: Finite Element model of a printing squeegee was created on the base of squeegee deformation experiments; squeegee is inserted into the stencil FEM model. 3.Stencil deformation by squeegee loading: Stencil is loaded at different locations by a printing squeegee; different sizes of underside supports were used; final FEM model of the stencil was created (including the squeegee) according to deformation results. 4.Testboard for stencil printing experiment: Testboard was designed according to BOSCH requirements; thick steps (with different distances from pads) were formed by selective electroplating. 5.Stencil printing experiment: Stencil printing experiment was carried out using testboards with different step thicknesses; the deposited paste height was measured and simulated; keepout area rule was set up on the basis of simulation results.

4. BOSCH - STENCIL_FEM 4/59 1. Stencil deformation by point loading

5. BOSCH - STENCIL_FEM 5/59 STENCIL DEFORMATION EXPERIMENT frame 58x58 cm stencil 50x50 cm underside support 30x30 cm 16 cm Stencil: - stainless steel - lasercut - thickness: 175 µm - ordered from DEK

6. BOSCH - STENCIL_FEM 6/59 STENCIL DEFORMATION EXPERIMENT free to move measuring probe fixed micrometer clock clock standstencilclock fixing m loading armclock stencilunderside support 365 mm 953 mm Clock: range: 0…1 mm accuracy: 10 µm Load: 2.26…7.5 kg / 22…131 N

7. BOSCH - STENCIL_FEM 7/59 STENCIL DEFORMATION RESULTS

8. BOSCH - STENCIL_FEM 8/59 FEM MODEL OF THE STENCIL FOIL FEM model has been created for both 16 cm and 30 cm underside support, to match to both experimental results. The material parameters obtained from Comsol library: Steel AISI 4340 – E: 205·10 9 Pa, ν: 0.28, ρ:7850 kg/m 3 stencil thickness: 175 µm

9. BOSCH - STENCIL_FEM 9/59 SIMULATION RESULTS measured: 2.84 µm/N, simulated: 3.03 µm/N measured: 1.95 µm/N, simulated: 2.11 µm/N

10. BOSCH - STENCIL_FEM 10/59 STENCIL DEFORMATION EXPERIMENT frame 58x58 cm stencil 50x50 cm underside support 30x30 cm 16 cm Stencil deformation experiment has been extended by underside supports of 10x10 cm and 5x5 cm.

11. BOSCH - STENCIL_FEM 11/59 STENCIL DEFORMATION RESULTS Stencil thickness: 125 µm

12. BOSCH - STENCIL_FEM 12/59 2. FEM model of a real squeegee

13. BOSCH - STENCIL_FEM 13/59 DEFORMATION OF A REAL SQUEEGEE The bending of the squeegee measured with the same loads as the stencil before to create the FEM model of the squeegee.

14. BOSCH - STENCIL_FEM 14/59 SQUEEGEE UNDER INVESTIGATION Length: 300 mm Blade: stainless steel Thickness: 200 µm 19 15 35

15. BOSCH - STENCIL_FEM 15/59 BOSCH - STENCIL_FEM SIMULATING SQUEEGEE DEFORMATION FEM parameters of the squeegee: Length: 300 mm Thickness: 200 µm Height: 15 mm Initial angle: 60 ° E : 195·10 9 Pa ν : 0.28 ρ : 7850 kg/m3

16. BOSCH - STENCIL_FEM 16/59 BOSCH - STENCIL_FEM SIMULATING SQUEEGEE DEFORMATION

17. BOSCH - STENCIL_FEM 17/59 3. Stencil deformation by squeegee loading

18. BOSCH - STENCIL_FEM 18/59 BOSCH - STENCIL_FEM MEASURING STENCIL DEFORMATION PUSHED WITH SQUEEGEE x y Squeegee length: 300 mm Stencil thickness: 125 µm Loads are the same as previous: 22…131 N Underside support: 31 cm, 20 cm, 10 cm

19. BOSCH - STENCIL_FEM 19/59 BOSCH - STENCIL_FEM FEM MODEL OF THE STENCIL WITH SQUEEGEE Stencil dimensions: real size – 580 mm x 580 mm x 125 µm Mesh: 1:1:100 (x:y:z) ratio for numerical accuracy, finer mesh size at pressure area (see figures) Squeegee is pressed from the top side, by uniform pressure Material properties: steel, E=195 GPa, Poisson’s Ratio: 0,28. Boundary conditions: surfaces inside the supported area can move and bend, other surfaces are fixed Squeegee location: y=0 mm Support size in example: 20 cm Width of support system: 30 mm Support size in example: 20x20 cm

20. BOSCH - STENCIL_FEM 20/59 BOSCH - STENCIL_FEM DEFORMATION IN X DIRECTION, 30 cm SUPPORT

21. BOSCH - STENCIL_FEM 21/59 BOSCH - STENCIL_FEM DEFORMATION IN X DIRECTION, 20 cm SUPPORT

22. BOSCH - STENCIL_FEM 22/59 BOSCH - STENCIL_FEM DEFORMATION IN X DIRECTION, 10 cm SUPPORT

23. BOSCH - STENCIL_FEM 23/59 BOSCH - STENCIL_FEM DEFORMATION IN Y DIRECTION, 30 cm SUPPORT

24. BOSCH - STENCIL_FEM 24/59 BOSCH - STENCIL_FEM DEFORMATION IN Y DIRECTION, 20 cm SUPPORT

25. BOSCH - STENCIL_FEM 25/59 BOSCH - STENCIL_FEM DEFORMATION IN Y DIRECTION, 10 cm SUPPORT

26. BOSCH - STENCIL_FEM 26/59 BOSCH - STENCIL_FEM MEASURING STENCIL DEFORMATION OF THE 150 µm TEST STENCIL Squeegee length: 300 mm Stencil thickness: 150 µm Loads are the same as previous: 22…131 N Point load and squeegee load is applied too Underside support: 20 cm

27. BOSCH - STENCIL_FEM 27/59 BOSCH - STENCIL_FEM DEFORMATION OF 150 µm STENCIL IN CASE OF POINT LOADINGS

28. BOSCH - STENCIL_FEM 28/59 BOSCH - STENCIL_FEM DEFORMATION OF 150 µm STENCIL IN CASE OF SQUEEGEE LODING AT 55 mm FROM CENTRE

29. BOSCH - STENCIL_FEM 29/59 BOSCH - STENCIL_FEM DEFORMATION OF 150 µm STENCIL IN CASE OF SQUEEGEE LODING AT 25 mm FROM CENTRE Conclusion: including apertures in simulation is not necessary

30. BOSCH - STENCIL_FEM 30/59 4. Testboard for stencil printing experiment

31. BOSCH - STENCIL_FEM 31/59 THE TESTPATTERN Step in different height from board to board: for example +20 µm, +40 µm, +60 µm The height of steps is formed by selective electroplating The clearance between the steps and the pads is varying from 300 µm to 5 mm Stencil aperture for paste deposition (0.5x0.5 mm), the paste transfer efficiency is not affected by Area Ratio, base thickness 35 µm Clear pad for reference thickness of paste measurement

32. BOSCH - STENCIL_FEM 32/59 THE TESTBOARD The testboard was designed according to Bosch requirements. Base thickness: contour and pads Higher steps by selective electroplating

33. BOSCH - STENCIL_FEM 33/59 THE TESTBOARD Nine pieces of testboard were made with immersion Ag finish; 3-3 of each step thicknesses: +20 µm, +40 µm, +60 µm.

34. BOSCH - STENCIL_FEM 34/59 MEASURING THE STEP THICKNESSES The thickness of the steps was measured with a Tencor Alphastep 500. Horizontal range: 2 mm Vertical range: 10 nm…300 µm Vertical resolution: 0.1 µm or 2.5 nm

35. BOSCH - STENCIL_FEM 35/59 MEASUREMENT POINTS narrow steps square wide steps

36. BOSCH - STENCIL_FEM 36/59 RESULTS OF ALPHASTEP MEASURING

37. BOSCH - STENCIL_FEM 37/59 5. Stencil printing experiment

38. BOSCH - STENCIL_FEM 38/59 BOSCH - STENCIL_FEM STENCIL PRINTING EXPERIMENT Printer model: DEK 248 Accuracy: (achievable) ±25μm Repeatability: ±10 μm Printing speed: 10-70 mm/s Squeegee force: 0-150 N Experimental settings: Printing speed: 30 mm/s, squeegee force: 92 N, blade length: 300 mm, separation speed: 6mm/s, 5 testboards were used for process setup. 1. Print 1 testboard -> print 1 fake board -> dry clean of stencil underside (repeated for 3 testboards) 2. Stencil direction / board direction was changed, stencil cleaned by wet wipe and with pressured air 3. Same run steps as No. 1. for another 3 boards

39. BOSCH - STENCIL_FEM 39/59 BOSCH - STENCIL_FEM TEST RUN Board ID.:Direction of printing Narrow step height [µm] Wide step height [µm] Square1 height [µm] Square2 height [µm] ID1Vertical27202550 ID2Horizontal29243018 ID5Vertical55325653 ID6Vertical925890120 ID7Horizontal68505869 ID8Horizontal755380 vertical printing horizontal printing narrow steps wide steps square2 square1

40. BOSCH - STENCIL_FEM 40/59 BOSCH - STENCIL_FEM MEASURING DEPOSITED PASTE HEIGHT solder paste solder pad step Measuring equipment: ASC-Visionmaster 150 Maximum sample height: 5.1 cm Resolution: 1.78 μm Maximum measurable height: 365 μm Field of view: 2.1x2.8 mm

41. BOSCH - STENCIL_FEM 41/59 BOSCH - STENCIL_FEM ID. 2.: STEPS ARE PARALLEL TO SQUEEGEE *Line is only for guide, not simulation result. Paste is higher than Cu step because if the stencil did not bend down to the pad, it lifted the paste during separation.

42. BOSCH - STENCIL_FEM 42/59 BOSCH - STENCIL_FEM ID. 7.: STEPS ARE PARALLEL TO SQUEEGEE *Line is only for guide, not simulation result. Paste is higher than Cu step because if the stencil did not bend down to the pad, it lifted the paste during separation.

43. BOSCH - STENCIL_FEM 43/59 BOSCH - STENCIL_FEM ID. 8.: STEPS ARE PARALLEL TO SQUEEGEE *Line is only for guide, not simulation result. Paste is higher than Cu step because if the stencil did not bend down to the pad, it lifted the paste during separation.

44. BOSCH - STENCIL_FEM 44/59 BOSCH - STENCIL_FEM ID. 1.: STEPS ARE PERPENDICULAR TO SQUEEGEE *Line is only for guide, not simulation result. Paste is higher than Cu step because if the stencil did not bend down to the pad, it lifted the paste during separation.

45. BOSCH - STENCIL_FEM 45/59 BOSCH - STENCIL_FEM ID. 5.: STEPS ARE PERPENDICULAR TO SQUEEGEE *Line is only for guide, not simulation result. Paste is higher than Cu step because if the stencil did not bend down to the pad, it lifted the paste during separation.

46. BOSCH - STENCIL_FEM 46/59 BOSCH - STENCIL_FEM ID. 6.: STEPS ARE PERPENDICULAR TO SQUEEGEE *Line is only for guide, not simulation result. Paste is higher than Cu step because if the stencil did not bend down to the pad, it lifted the paste during separation.

47. BOSCH - STENCIL_FEM 47/59 BOSCH - STENCIL_FEM STENCIL DEFORMATION OF NARROW STEPS Conclusion: if steps are perpendicular to the squeegee, the deposited paste has higher height with higher deviation, and the stencil can bend less.

48. BOSCH - STENCIL_FEM 48/59 BOSCH - STENCIL_FEM MEASURING THE DEPOSITED PASTE AREA Paste area was measured when the steps were perpendicular to printing direction The results were averaged from the pads outlined by the red rectangle

49. BOSCH - STENCIL_FEM 49/59 BOSCH - STENCIL_FEM MEASURING THE DEPOSITED PASTE AREA Left: ID. 1 - no step Right: ID. 6 – 0.5 mm step distance

50. BOSCH - STENCIL_FEM 50/59 BOSCH - STENCIL_FEM 3D SIMULATIONS FOR DIFFERENT DIRECTION OF PRINTING Simulations showed basically different printing process depending on the printing direction. The bending of the stencil can be senn in the figures above.

51. BOSCH - STENCIL_FEM 51/59 BOSCH - STENCIL_FEM DIFFERENT PLOTS IN DIFFERENT CASES -The simulation data for slides (52-57) were extracted from 3D simulations like on the previous slide -BUT there are two different types of simulation plots: -If the printing direction is parallel to the row of pads, the simulation doesn’t show a cross-setion of the bending stencil. -If the printing direction is perpendicular to the row of pads, the simulation data is a cross-section of the stencil. Example: cross-section data from the left image on the previous slide

52. BOSCH - STENCIL_FEM 52/59 BOSCH - STENCIL_FEM ID. 2.: PADS NEAR TO LARGE Cu SQUARE

53. BOSCH - STENCIL_FEM 53/59 BOSCH - STENCIL_FEM ID. 7.: PADS NEAR TO LARGE Cu SQUARE

54. BOSCH - STENCIL_FEM 54/59 BOSCH - STENCIL_FEM ID. 8.: PADS NEAR TO LARGE Cu SQUARE

55. BOSCH - STENCIL_FEM 55/59 BOSCH - STENCIL_FEM ID. 1.: PADS NEAR TO LARGE Cu SQUARE

56. BOSCH - STENCIL_FEM 56/59 BOSCH - STENCIL_FEM ID. 5.: PADS NEAR TO LARGE Cu SQUARE

57. BOSCH - STENCIL_FEM 57/59 BOSCH - STENCIL_FEM ID. 6.: PADS NEAR TO LARGE Cu SQUARE

58. BOSCH - STENCIL_FEM 58/59 BOSCH - STENCIL_FEM SIMULATING THE KEEPOUT AREA In the simulations 92 N squeegee force and perpendicular steps (as the worst case) were used.

59. BOSCH - STENCIL_FEM 59/59 SUMMARY -FEM model of the stencil is created including a real squeegee -Printing experiments were carried out. -Steps, perpendicular to squeegee line cause higher paste deposit with higher deviation, and the stencil can bend less. -According to the simulations, the minimum recommended keepout area is 1.6*step_height*stencil_foil_thickness. -This recommendation can be used for step stencils as well (for PIP technology or for mixed-pitch applications) instead of the IPC-7525 standard ’36*step_height’ rule.