ECE535_DesignProj_Report_ABowers&HDhrimaj (1)

Anne Bowers ECE/MAE 535 July 14, 2014 Helion Dhrimaj Summer 2014

NORTH CAROLINA STATE UNIVERSITY Design of Electromechanical Systems

Semester Design Project

Electro-‐Permanent Magnet Clamp

ABSTRACT Manufacturing processes that involve ‘material removal’ (such as milling, drilling, etc.) depend upon a secure attachment of the work piece to the machine table in order for the work to be completed safely and with the desired precision. Traditionally a mechanical clamp would be used to secure the work piece, but this process can become tedious and time consuming. In an effort to improve upon these two issues, an electro-‐magnetic solution was developed. An electromagnet would be turned on to secure the work piece to the table, due to reluctance forces, and then would be switched off when the process was finished. This solution was able to provide the desired faster and easier set up, but now had the additional risk of catastrophic failure during a power interruption, and was expensive to operate due to high power consumption. The electro-‐permanent magnet clamp is a new design that strives to maintain the fast and easy setup provided by the electromagnetic clamp, while making use of rare earth magnets to minimize power consumption. INTRODUCTION This project will focus on the development of an optimized design for an electro-‐permanent magnet clamp. The basic design for the electro-‐permanent magnet clamp is to have two different kinds of rare earth magnets distributed within the workbench. NdFeB magnets are used to provide a strong flux density that is undisturbed regardless of being in the on or off state. AlNiCo magnets, which have a lower coercivity, are then used to direct the flux to either stay within the workbench, or through the work piece. The different states of the clamp, ‘on’ or ‘off’, are manipulated by placing the AlNiCo magnets within a wire coil that can reverse the polarity of the AlNiCo by pulsing current through the coil in either direction. This results in a clamp that is either ‘on’ or ‘off’ without the need to constantly run electricity to sustain an electro-‐magnet. Figure 1 shows the arrangement and polarity of the permanent magnets within the workbench when the clamp is in the ‘off’ state. There are three NdFeB magnets that are arranged so that their polarities are oriented horizontally. There are two AlNiCo magnets placed within coils of wire with their polarities oriented vertically. The arrows in the figure indicate the path of magnetic flux, and show how all flux is ideally contained within the workbench. This leaves the work piece free to move around for easy adjustment. Figure 2 shows that when the AlNiCo magnets are pulsed with sufficient current through their surrounding coils, a strong enough magnetic field can be generated to reverse their polarity. This reversal in polarity now directs flux through the work piece, creating a reluctance force that holds the work piece in place for processing. This is the


clamp ‘on’ state, and it does not require sustained electrical power to maintain the clamping force, making it both safer and more efficient than the electromagnet clamp.

Figure 1. Cutaway of Electro-‐Permanent Magnet Clamp in the ‘Off” State

Figure 2. Cutaway of Electro-‐Permanent Magnet Clamp in the ‘On’ State

The design optimizations for this clamp will focus on minimization of the overall size and weight of the electro-‐permanent magnet clamp (EMPC) and required permanent magnetic materials, NdFeB and AlNiCo, while still maintaining a minimum of 500lbs of vertical reluctance force while in the ‘on’ state. Minimization of these two parameters will reduce the overall cost to produce the clamps. The design will also seek to maximize the vertical reluctance force as a secondary goal to minimizing the material costs. METHODS AND MATERIALS The materials that will be used for this clamp are AlNiCo permanent magnets, NdFeB permanent magnets, 1025 steel, magnet wire, and a 230V source. The grades and volume of the permanent magnets will be part of the design optimization. The size of magnet wire to be used will need to be determined based on the number of turns required to generate the magnetic pulse that flips the AlNiCo polarization, and the amount of current


that can be sustained within the wire. Tables 1-‐3 summarize the material properties for the available permanent magnet materials and 1020 Steel. 1020 steel is being considered because it is the closest grade to 1025 that is available for simulation.

Table 1. Summary of Properties of Available NdFeB Materials Available NdFeB Materials

Grade Remanence Flux Density Coercive Max. Energy Density Br HcB HcJ (BH)max

mT G kA/m Oe kA/m mT G kA/m N10[3] 690 6,900 424 5,300 760 9,500 90-‐100 7.3-‐8.1 N38[1] 1,220 12,200 899 11,300 955 12,000 287 36 N40[1] 1,250 12,500 907 11,400 955 12,000 302 38 N42[1] 1,280 12,800 915 11,500 955 12,000 318 40 N52[1] 1,430 14,300 796 10,000 876 11,000 398 50

Table 2. Summary of Properties of Available AlNiCo Materials Available AlNiCo Materials[4]

Grade Remanence Flux Density

Coercive Max. Energy Density

Operating Temp

MMPA

Br HcB HcJ (BH)max Tw Max kGs mT kOe kA/m kOe kA/m MGOe kJ/m3 C

LNG34 11 1100 0.63 50 0.65 52 4.25 34 525 AlNiCo5 LNG37 11.8 1180 0.61 49 0.64 51 4.63 37 525 AlNiCo5 LNG40 12 1200 0.63 50 0.65 52 5 40 525 AlNiCo5 LNG44 12.5 1250 0.65 52 0.68 54 5.5 44 525 AlNiCo5 LNG52 13 1300 0.7 56 0.73 58 6.5 52 525 AlNiCo

5DG LNG60 13.5 1350 0.73 58 0.75 60 7.5 60 525 AlNiCo

5-‐7

Table 3. Summary of Properties of 1020 Steel AISI 1020 Steel Characteristics

Density 0.2839 lb/in3 Tensile Strength 63800 psi Relative Permeability (0.3440 T)[2] 1496 Relative Permeability (0.9600 T)[2] 444 Relative Permeability (1.4700 T)[2] 97 Relative Permeability (1.6150 T)[2] 55

The data in Table 1 shows that the increase in grade for NdFeB corresponds to an increase in BHmax, the maximum energy product. It is expected that higher grades of NdFeB will require less material to make a clamp with sufficient holding force, but depending on


the market value for different grades, the highest and lowest grade may not prove to be the most economical choice. The data in Table 2 indicates that the lower grades of AlNiCo have a lower coercivity, and so it should take less current in the coils to flip the magnet poles making it a more efficient clamp to operate. The lower grades will also require more material because their maximum energy density is lower. The magnetic flux that is responsible for generating the holding force while the clamp is in the ‘on’ state must also be fully contained in the clamp during the ‘off’ state in order to make it easy to position and adjust the work piece before clamping it down for work. In order to prevent any flux from travelling through the work piece while the clamp is in the ‘off’ state, the flux supplied by the AlNiCos must equal the flux supplied by NdFeBs. This can be calculated using equation 1, and confirmed by simulation using FEMM.

𝐵!"#$%& ∗ 𝐴!"#$%& = 𝐵!"#$% ∗ 𝐴!"#$% Eq. (1) The data in Table 3 includes the relative permittivity values for steel at different flux density values. This is important because it indicates when the steel in the clamp and work piece are reaching their saturation point. If the steel becomes saturated then energy is being wasted, and a lower grade of magnet that supplies less flux density will be able to achieve the same holding force, presumably for less cost. Table 3 shows that steel becomes saturated between 1.4 T and 1.6 T. Any designs that result in points of flux density greater than this range should not be considered. The last parameter is the ability to switch the clamp ‘on’ and ‘off’ using the coil bobbins. The minimum current pulse required in the coil bobbins can be predicted using Ampere’s Law. Ampere’s Law relates the current enclosed by a closed contour integral to the magnetic field. Equations 2, 3, and 4 show Ampere’s Law and how it will be used to determine the minimum current for turning the clamp on and off.

𝐻�𝑑𝑙!! = 𝐼!"#! Eq. (2)

𝐻 ∗ 𝐿 = 𝑁 ∗ 𝐼 Eq. (3)

!∗!!= 𝑁𝐼 → 𝐼 = !∗!

!∗! Eq. (4)

Given Equation 4 we can determine the necessary current to turn the clamp on and off. This relationship shows that the necessary current is directly related to the strength of the magnetic flux of the AlNiCo magnets, and inversely proportional to the number of turns wire. The length of wire will be limited by the coil bobbin dimensions and available space, and the permeability will be determined by the grade of AlNiCo chosen in the final design. The resistance of the wire is also a factor in determining how much current can be pulsed with a 230V source. The physical properties of several common gauges of magnet wire are summarized in Table 4.


Table 4. Physical Properties of Different Magnet Wire Sizes Magnet Wire Specifications Wire Gauge (AWG)

Resistance (Ω/1000ft)

Total Resistance for 97 Turns (Ω)

Maximum Current w/ 230 V Source

Cross-‐sectional Area for 97 Turns (in2)

24 25.67 2.27 101.3 0.031 22 16.14 1.43 160.8 0.049 20 10.15 0.897 256.4 0.078 18 6.385 0.565 407.1 0.124 16 4.016 0.355 647.9 0.197 Using the FEMM software introduced in the course, and the provided general layout, a model was constructed. Figure 3 shows this layout, including the materials that were used in the simulation. 1020 steel was used for the simulation because it was the closest available material in the material library. The air gap is 0.1 mm wide, and the model shown has a depth of 2.6cm, which is midway through the clamp. All forces calculated by FEMM will need to be doubled to account for the other half of the clamp.

Figure 3. Cross-‐sectional layout of Electro-‐Permanent Magnet Clamp

RESULTS The initial magnetics simulation results, using NdFeB 37 and AlNiCo 5 were able to achieve an attractive force between the work piece and the body of 6034 N, which is roughly 1356 lbs. This design meets the minimum force specifications, but can be improved. The magnetic flux was concentrated in the two poles of the workbench at nearly 2 Tesla. Given that steel reaches saturation between 1.4 and 1.6 Tesla[2], energy was being wasted in this saturated condition. The clamp was redesigned to help minimize this inefficiency in each different magnet grade combination, which is summarized in Table 6. The minimum clamp dimensions that were found to ensure a reasonable path for flux without unnecessary bulk are summarized in Table 5. While the overall size of the clamp is not large in dimensions, it is ultimately a heavy piece of equipment, totaling around 62 pounds. The varying height of the AlNiCo magnets for each design made it difficult to predict exactly how much steel would be in the pole without knowing the final client design choice, so the total weight is a conservative maximum based on the presence of no AlNiCo. Any final clamp will have slightly less steel.


Table 5. Clamp Steel Weight

Clamp Steel Weight Based 1025 Steel Density Clamp Piece Volume (mm3) Volume (in3) Density (lb/in3) Weight (lb) Pole* 245000 14.951 0.2839 4.24 Base 1485000 90.620 0.2839 25.73 Sides 1862000 113.63 0.2839 32.26 Total 3592000 219.201 -‐ 62.23

Table 6. Summary of Minimum Required Dimensions and Cost for Different Magnet Grade

Combinations Magnet Combination

Force in ON (N)/(lbs)

Force in OFF (N)/(lbs)

Max. Satu-‐ ration (T)

NdFeB Dimensions (mm)

AlNiCo Dimensions (mm)

NdFeB Price ($)

AlNiCo Price ($)

I (A)

NdFeB10[8-‐11]/ AlNiCo 5[6]

4552.99/1023

~0/0 1.41 (60 x 50 x 36)

(47 x 46 x 17)

324 103.21 66.6

NdFeB10[8-‐11]/ AlNiCo 5[6]

6432.22/1446

~0/0 1.71 (60 x 50 x 44)

(47 x 46 x 45)

396 274.99 188.9

NdFeB37[8]/ AlNiCo6[5][6][7]

2422.7/ 544

34/7.6 1.293 (60 x 50 x 17)

(47 x 46 x 10)

699.50 80.57 10

NdFeB37[8]/ AlNiCo6[5][6][7]

6773.06/1522

34/7.6 1.747 (60 x 50 x 17)

(47 x 46 x 30)

699.50 241.72 100

NdFeB40[9]/ AlNiCo6[5][6][7]

2676.94/601

36.38/ 8.1

1.36 (60 x 50 x 17)

(47 x 46 x 10)

141.984 80.57 10

NdFeB40[9]/ AlNiCo6[5][6][7]

6853.24/1540

36.37/ 8.1

1.814 (60 x 50 x 17)

(47 x 46 x 30)

141.98 241.72 100

NdFeB52[10]/ AlNiCo6[5][6][7]

2488.2/ 559

32.157/7.2

1.374 (60 x 50 x 17)

(47 x 46 x 10)

199.18 80.57 10

NdFeB52[10]/ AlNiCo6[5][6][7]

6712/ 1508

36.637/8.2

1.84 (60 x 50 x 17)

(47 x 46 x 30)

199.18 241.72 100

NdFeB37[8]/ AlNiCo8[5]

2378.9/ 535

47.178/10.6

1.351 (60 x 50 x 17)

(47 x 46 x 12.5)

690.50 75.72 24


6669.97/1500

44.62/ 10.0

1.813 (60 x 50 x 17)

(47 x 46 x 32.5)

690.50 196.89 150


2463.62/ 554

57.72/ 13.0

1.366 (60 x 50 x 17)

(47 x 46 x 12.5)

141.984 75.72 24


6756.47/1519

54.659/12.3

1.837 (60 x 50 x 17)

(47 x 46 x 32.5)

141.984 196.89 150


2565.4/ 577

127/ 28.6

1.307 (60 x 50 x 17)

(47 x 46 x 12.5)

199.18 75.72 22


6794.49/1527

124.4/28.0

1.839 (60 x 50 x 17)

(47 x 46 x 35)

199.18 212.04 150


Figures 4-‐6 show results of simulations made with NdFeB 10 and AlNiCo 5 magnets. Simulations of all other magnet combinations yielded similar results. Figure 4 shows the clamp in the ‘on’ position. The maximum flux density is concentrated in the poles and the work piece but does not exceed 1.41T so minimal amounts of energy are being wasted. Figure 5 shows the same clamp in the off state with a measured 0.05lbs of off force. That amount of force should be unnoticeable to anyone responsible for positioning the work piece. The last figure, figure 6, shows the clamp in the midst of switching. The lack of flux lines through the AlNiCo magnets and coils shows that the current in the coil is at least strong enough to balance the AlNiCo flux. Any current increase past that point will cause the AlNiCo magnet poles to flip.

Figure 4. Clamp with Minimum PM Materials N10 and AlNiCo 5 in On Position with 2046lbs

of holding force

Figure 5. Clamp with Minimum PM Materials N10 and AlNiCo 5 in Off Position with -‐0.05lbs

of Holding Force


Figure 6. Clamp with 66.6 A of Current During Switching from On to Off

CONCLUSIONS The NdFeb10/AlNiCo5 combination achieved the required force with a force of nearly zero in the “OFF” state. For our clamp design, this proved unobtainable with the higher-‐grade magnets. As evident in the table above, all of the higher grade magnetic combinations experienced some force in the air gap during the “OFF” state. With the NdFeB37, 40, and 50, the ideal dimensions ranged from (60 mm x 50mm x 15mm) to (60 mm x 50mm x 22.5 mm). This NdFeB37/40/50 dimensional range supplied enough magnetic flux density to achieve 500 lbs. and 1500 lbs. of force. At the same time it minimized the magnetic saturation as well as the force in the OFF state. For the sake of time, we chose to leave the NdFeB magnets at a constant (60mm x 50mm x 17mm). The only variable left to modify was the AlNiCo dimension. According to the reluctance force in the “ON” state, the AlNiCo5 length ranged from (17 mm -‐ 45 mm), the AlNiCo6 length ranged from (10 mm-‐ 30 mm), and the AlNiCo8 length ranged from (12.5mm to 35 mm). Combinations of higher magnetic grades, NdFeB37 and above, as well as AlNiCo6 and above, provided sufficient force for the clamp. On the negative side, these aforementioned magnets experienced high levels of saturation when designed for the 1500 lbs. of force. Furthermore, an air gap force was measured in the FEMM analysis even in the OFF state. This is more substantial in the NdFeB52/AlNiCo8 combination, where we measured a force close to 130 N in “OFF” state. The maximum cross-‐sectional area that is available for the coil is 0.558in2 so the clamp dimensions would accommodate any of the wire sizes in table 4. The larger size wire would be more desirable for the final product because it is better able to handle the higher current pulses and the heat that is generated during the pulse. Overall it would be difficult to choose an ideal magnet design. The NdFeB10/AlNiCo5 combination has a high price due to the amount of material being used. At the same time it experiences almost no force in the “Off” state and relatively low saturation levels, even for the 1500 lbs. force requirement. Alternatively, the higher grades of magnet tend to require less overall material. This equates to slightly lower costs but at the same time higher levels of saturation result, and a force is present in the “Off” state. Considering the alternatives we were able to develop, and our ability to apply FEMM as a tool in the design process, this project has been a good opportunity to demonstrate the material we have learned throughout the semester.


REFERENCES [1] NdFeB Specialists E-‐Magnets, Berkhamsted, Herdforshire, UK (2014). “Grades of Neodymium” (Online) Accessed 29 July, 2014. http://www.ndfeb-‐info.com/neodymium_grades.aspx [2] Field Precision LLC., Albuquerque, New Mexico, USA (2014). “Saturation Curves for Soft Magnetic Materials” (Online) Accessed 29 July, 2014. http://www.fieldp.com/magneticproperties.html [3] Magnetic Materials & Components, Hauppage, New York, USA (2014). “Compression Bonded Magnets” (Online) Accessed 29 July, 2014. http://www.mmcmagnetics.com/ourproducts/Bonded/Compression_Bonded.htm [4] X-‐Mag, Inc., Hangzhou, Zhejiang, China. (2014). "Magnetic Characteristics of AlNiCo Magnets". [5] Arnold Magnetic Technologies Corporation, Rochester, New York, USA (2014). “Alnico 8 Magnets” (Online) Accessed 29 July, 2014. http://buyonline.arnoldmagnetics.com/p-‐11-‐alnico-‐8-‐magnets.aspx [6] Arnold Magnetic Technologies Corporation, Rochester, New York, USA (2014). “Alnico 5 Magnets” (Online) Accessed 29 July, 2014. http://buyonline.arnoldmagnetics.com/p-‐12-‐alnico-‐4-‐5-‐magnets.aspx [7] Total Magnetic Solutions, Culver City, California, USA (2014). “Properties of Alnico Magnets” (Online) Accessed 29 July, 2014. http://www.magnetsales.com/alnico/Alprops.htm [8] Magnet Shop, Culver City, California, USA (2014). “Neodymium Block Magnets” (Online) Accessed 29 July, 2014. http://magnetsales.thomasnet.com/viewitems/block-‐magnets/large-‐neodymium-‐rectangle-‐magnets?&forward=1 [9] CMS Magnetics, Garland, Texas, USA (2014). “Neodymium Magnets” (Online) Accessed 29 July, 2014. http://www.magnet4sale.com/10-‐pc-‐n40-‐neodymium-‐magnets-‐4-‐10x2-‐10x1-‐10-‐ndfeb-‐rare-‐earth-‐magnets/ [10] K & J Magnetics, Inc., Pipersville, Pennsylvania, USA (2014). “Neodymium Block Magnets” (Online) Accessed 29 July, 2014. http://www.kjmagnetics.com/products.asp?cat=11 [11] Integrated Magnetics, Culver City, California, USA (2014). “Neodymium Iron Boron Magnets-‐ General Information” (Online) Accessed 29 July, 2014 http://www.intemag.com/NdFeB.html

ECE535_DesignProj_Report_ABowers&HDhrimaj (1)

Documents

permanent magnets

alnico magnets

electromagnet clamp

electromagnetic clamp

mechanical clamp

ndfeb magnets

work piece free

electromagnetic solution