Research on High-Speed Vibration and Structure Optimization of Multiwire Sawing Machine

Abstract

To improve the sawing stability and slicing quality of an existing multi-wire saw under severe vibrations using a high wire speed, its structure is optimized to improve its structural stiffness, so that it can still maintain good operation stability under higher wire speed. First, a finite element simulation model is established based on the shape, structural and material characteristics of the multi-wire saw, and then a transient dynamic simulation calculation is carried out using finite element software to obtain the vibration characteristics of the machine, which is verified with the measured data. The results show that the error between the calculated and the measured value is less than 11.46%, which verifies the accuracy of the simulation. On this basis, a multi-objective optimization design method based on the response surface is used to optimize the structure of the whole machine, and the optimized model is simulated again to evaluate its optimization effect. The results show that the amplitude of the optimized machine at a 2400 m/min wire speed is equivalent to that of the original machine at a 1500 m/min wire speed, and the wire speed is increased by 60% year-on-year with the same operation stability, which can guide research on high-speed multi-wire saws and machine tool vibration control strategies.

1. Introduction

A multi-wire sawing machine is a high-precision machine that cuts materials via friction using the sawing edge attached to the steel wire moving at a high speed [1]. Compared with a traditional inner circle slicer, the multi-wire sawing machine has high efficiency, high capacity, and high precision for slicing semiconductor materials and is the most widely used wafer sawing technology at present [2,3]. With the rapid development of the photovoltaic and semiconductor industries, the existing sawing efficiency does not meet the production needs of enterprises; to improve the sawing speed and efficiency of the material, it is necessary to improve its alignment speed, so that it will be able to withstand high speeds [4,5]. Increasing the alignment speed can reduce the sawing time, material consumption, production costs, etc., but with the increasing speed of the alignment the operational stability requirements are also increasingly high [6,7].

After research, although some manufacturers launched a multi-wire sawing machine with a 1800 m/min line speed, the actual sawing stability of the machine tool is limited; when it runs at a line speed of 1800 m/min, the violent vibrations of the machine tool reduce the slicing quality, indicating that its existing structure does not meet the requirements of the use of a high line speed. Therefore, the study of the dynamic characteristics and operational stability of multi-wire sawing machines at high wire speeds is of great importance to the slicing industry [8,9]. Similarly, by optimizing the structural configuration of the machine tool, the material is more rationally distributed and effectively utilized to improve its performance, which is also an important goal pursued by the designer [10,11].

There are minimal published results on high-wire-speed multi-wire sawing machine research, but there are many research results on the dynamic and static characteristics and optimal design of multi-wire sawing machines. Liu [12] performed finite element modeling and static mechanical analysis of a multi-wire sawing machine. Yanbo et al. [13] performed a dynamic mechanical analysis of a multi-wire sawing machine and compared the modalities of each order. Meng [14] and Shenglin [15] studied and analyzed the static and dynamic characteristics of a machine tool and optimized its structure. Choi et al. [16] used a genetic algorithm for the topology optimization of the machine structure. Kolar et al. [17], and Lee et al. [18] designed a lightweight machine column and spindle system using topology optimization, respectively. Dun et al. [19] improved the dynamic and static performance by optimizing the spindle support tripod, while reducing the mass. However, most of the above static and dynamic analyses are based on the modal analysis of only a certain part of the structural optimization, and not for the comprehensive dynamic simulation and optimization of the whole machine, which would facilitate our understanding of the operating stability of multi-wire sawing machines at high wire speeds.

In this paper, the finite element dynamics simulation of a model of a multi-wire sawing machine is carried out to obtain its vibration characteristics, and the validity and accuracy of the simulation are verified through experimental measurements. The results of the dynamics calculation are then used as the basis for the multi-objective structural optimization design of the machine using the response surface method, and the optimization effect is evaluated using dynamics verification calculations. Finally, the structural design of the machine is proposed to be improved for the development of a multi-wire sawing machine with a high wire speed.

2. Materials and Methods

2.1. Principle Analysis of Vibration Characteristics

A multi-wire sawing machine spindle mounted on a groove wheel with hundreds of diamonds wire saw is shown in Figure 1. The tension of many wire saws superimposed together will make the hollow structure of the slotted wheel deform, and the rotating parts in the spindle assembly are structurally complex. When the spindle is driven by the motor rotating at a high speed, it will cause the machine to self-excitedly vibrate, and the higher the speed is, the greater the excitation force generated is, and when the amplitude of vibration is greater the area above the main frame of the material platform will also vibrate, and the sawing stability will be reduced. This is one of the main reasons why the surface quality of the slices deteriorates when multi-wire sawing machines are run at higher line speeds.

Dynamic stiffness is the main indicator of machine tool vibration resistance; the dynamic stiffness of a machine tool can be expressed by using the following mathematical expression:

$$K=\frac{F}{A}$$

(1)

where K indicates the dynamic stiffness of the machine tool, F indicates that the machine tool is subject to the excitation force, and A indicates the size of the machine amplitude.

For a multi-wire sawing machine at a certain alignment speed, the machine is subject to a certain maximum excitation force F. In this state, if the dynamic stiffness K of the machine can be improved, the vibration amplitude A of the machine can be reduced accordingly, thus improving the operational stability of the multi-wire sawing machine and improving the surface quality of the slices.

2.2. Whole Machine Transient Dynamics Simulation

2.2.1. Finite Element Modeling

A finite element model of the whole frame of a multi-wire sawing machine was established, as shown in Figure 2. The authors considered the mesh division, boundary conditions, importance of the simulation calculation, and processing and manufacturing processes. The main body of the structural support is retained, the micro-geometric features are cleaned, and small structures, such as individual holes and rounded corners, are removed. Some additional non-structural supporting parts, such as the cutting fluid spray, liquid drain, lubrication device, and complex geometric mutation features, were simplified and removed, which made the structure of the finite element model smooth and regular, which is convenient for subsequent mesh division and calculation accuracy control.

2.2.2. Parameter and Boundary Condition Settings

The materials and basic parameters used for each part of the machine are shown in

Table 1. Material performance parameters.

Materials	Density /(kg·m−3)	Young’s Modulus /MPa	Poisson’s Ratio	Name of Parts
QT-450-10	7300	1.9 × 105	0.27	End caps, main frame.
38CrMoAlA	7830	2.06 × 105	0.29	Spindle, groove wheel steel core.
HT300	7300	1.43 × 105	0.27	Bottom bracket, subframe, electric control cabinet.
Polyurethane	1200	896	0.41	Grooved wheel jacket.

.

After analysis and calculation, the load on the whole machine of this multiline sawing machine mainly includes the following:

• A pre-load force of 58400 N for the long bolt on the outside of the swim end, where the torque of the long bolt is 250–280 N·m, taking the larger value of the upper boundary.
• The wire tension of the diamond wire on the grooved wheel is 50 kN, the maximum tension of which is calculated according to the commonly used wire type and tension meter and rounded upwards.
• Dual spindle motor speeds, where a slotted wheel diameter of 180 mm corresponds to different speeds for different alignment speeds.
• Mass of the material feed platform section: 700 kg.
• The take-up and payoff wheel mass is 70 kg.
• The take-up and payoff motor mass is 206 kg.
• Tension control motor with a mass of 46 kg.
• The mass of the motor and screw parts of the discharge line is 33 kg.
• The mass of the electrical components in the electric control cabinet is 300 kg.
• The self-weight of the machine frame, etc.

Mass point replacement is applied to the material feed platform, motor, and other parts above the main frame. The position and size of each center of mass are determined, and their corresponding actions on the frame are established. The MPC treatment was applied to the frame part, a fixed support treatment was applied to the bottom surface of the tab of the bottom bracket, and the contact was set according to the actual situation.

Due to there being a large number of load types, the loads are applied in steps, and separate substeps are set for each load, and the substeps for different working conditions are reasonably tuned to reduce the number of dichotomies in the calculation process and accelerate the convergence speed in the calculation. Hybrid area meshing with advanced nodes is used for the whole frame, the mapping type is a hexahedron, and the free mesh type uses a hexahedron as the core. The spindle part is meshed separately, and its size is kept at 25 mm, The boundary conditions for finite element analysis are applied as shown in Figure 3. The arrows shown in Figure 3 represent the directions of the forces and torques applied as boundary conditions in the multi-wire sawing machine

During the operation of the multi-wire sawing machine, the spindle motor moves in a certain cycle pattern, which makes the wire saw reciprocate and cut the material. Since the time variation is not considered in static devices for such complex conditions, although the dynamics calculation is tens or even hundreds of times more time-consuming than that for static devices, it is necessary to perform dynamic finite element simulations to understand the vibration characteristics more accurately.

Since the transient dynamics simulation is time-consuming, the calculation time is simplified by taking one acceleration/deceleration cycle of the motor operating as a calculation time length, and the vibration characteristics of the machine tool at different linear speeds are simulated using dynamic finite element analysis.

3. Results

3.1. Finite Element Simulation Result

After completing the simulation, data post-processing is performed using the corresponding simulation results. The vibration characteristics of the whole frame when running at a 2400 m/min line speed are shown in Figure 4. When running at a 1200 m/min line speed, the variation in the deformation of the main frame over one calculated time length is shown in Figure 5.

The difference in the deformation results is the vibration amplitude; the maximum value of the amplitude of each part of the machine tool is extracted, and data post-processing is carried out to obtain the amplitude of each part of the multi-wire sawing machine at different line speeds, as shown in Figure 6.

From Figure 6, it can be seen that the amplitude increases non-linearly with the line speed, especially when the line speed is greater than 1200 m/min. The slope of the amplitude change increases sharply, and from the overall point of view, the amplitude of the subframe is greater than the other parts, and the amplitude of the bottom bracket part is smaller than the other parts.

The first-order modal vibration pattern of the whole machine is shown in Figure 7 below, which was obtained after the analysis of the dynamic characteristics of the multi-wire sawing machine using finite element software.

For the analysis of the multi-wire sawing machine, the structure is stable when its intrinsic frequency is constant, and the higher the speed of the spindle components is, the greater the frequency of excitation is, and the more likely it is to resonate, which has a hugely damaging effect on the sawing of materials. As can be seen from the figure, the first-order modal value of the whole frame is 63.62 Hz, and the excitation frequency is 63 Hz at a line speed of 2100 m/min, for which time resonance may occur. Therefore, to avoid resonance, we must consider an appropriate reduction in the speed of the multi-wire sawing machine, thereby reducing its excitation frequency. In the case of a certain line speed, when the diameter of the slot wheel is larger, the corresponding motor speed will be lower.

3.2. Vibration Measurement Test

Through the transient dynamic finite element simulation of the whole machine, the vibration characteristics of the multi-wire sawing machine at different line speeds were obtained. Vibration measurement tests were performed to verify the validity and accuracy of the simulation and to provide a reference basis for subsequent structural optimization and machine upgrades.

According to the finite element simulation results, the maximum deformation positions of each component of the whole multi-wire sawing machine were detected, as shown in Figure 8, which provides a reference for the location arrangement of the sensors during the experimental test.

Displacement sensors were selectively distributed on each part so that the sampling points are evenly distributed, and the locations of the largest and smallest displacements were accounted for in four separate vibration tests on four parts of the machine, corresponding to the maximum deformation of the sensor site layout, as shown in Figure 9. To avoid accidentally obtaining the measurement data, six acceleration sensors of the same model were used to take vibration measurements at the same time.

The test variable studied in the vibration measurement test on the whole multi-wire sawing machine is the line speed because the actual machine can only run at a maximum line speed of 1800 m/min; therefore, the line speeds selected and used during the test were 600, 900, 1200, 1500, and 1800 m/min.

3.3. Experimental Results

Finally, after post-processing the test data, the maximum amplitudes of each part of the machine tool at different line speeds were obtained, as shown in Figure 10.

Compared with the simulation measurements, the experimental measurements of the bottom bracket and subframe are significantly larger. The reason for this is that the bottom bracket was set to fixed support in the finite element simulation calculation, which limits the displacement of the bottom bracket to some extent. The boundary conditions of the subframe winding wheel motor are simplified, resulting in the excitation force on the subframe being smaller than the actual one, so the calculated amplitude value of the bottom bracket and subframe is small, which also leads to the amplitude of the other components being smaller than the test measurement. The above settings and processing of the bottom bracket and subframe make the finite element dynamics simulations easy to converge and calculate the results for.

To ensure the sawing stability of a multi-wire sawing machine, the most important part is the main frame. The main frame is located above a cloth feed platform, and the winding diamond wire saw slot wheel is located in the main frame part. If the amplitude of the main frame is large, then the slicing quality is relatively poor. Therefore, to describe the sawing stability of the multi-wire sawing machine more accurately, the amplitude calculation accuracy of the main frame is the main measurement of the validity and accuracy of the simulation.

In comparison, the amplitudes of the test frame and electric control cabinet are basically consistent with the simulation results. Considering the simplified simulation model process, the amplitude variation trend of the finite element calculation is the same as that of the field test, and the error is smaller; both are within 11.46%. When the line speed is 1800 m/min, the experimental and simulation results show that the amplitude is about 100 μm. Therefore, the simulation can accurately describe the variation in amplitude with linear velocity, the error between the electronic control cabinet and the main frame under different line speeds is shown in Figure 11.

The above analysis shows that a 1200 m/min line speed is the amplitude inflection point of the multi-wire sawing machine, and after this, the amplitude is significantly increased, and the simulation and test results are consistent with the trend of amplitude changes. These are also consistent with those in the research on industrial field use conditions. In this research, the machine site generally uses a line speed of 1200 m/min and a higher line speed for slicing; its surface quality is also reduced, and it cannot meet the production requirements.

It is determined that the abovementioned finite element simulation calculation has a certain validity and accuracy, which can provide a reference basis for the subsequent structural optimization. Therefore, the next structural optimization is carried out to maintain good operational stability even at higher line speeds.

3.4. Structural Optimization and Dynamics Verification

3.4.1. Overall Solution Analysis for Structural Optimization

To optimize the structure of the multi-wire sawing machine, we must first determine its optimization target, and we can only optimize the structure of the whole machine after this.

In the optimization process, first, parametric correlation analysis is performed on the model dimensions of the whole machine to determine the dimensional variables; second, numerous dimensional variables are screened using the sensitivity screening method to obtain the parameter variables with a large influence factor on the optimization target. Then, a response surface is constructed with the screened variables to obtain the correspondence between the parameter variables and the optimization target. Finally, based on the constructed response surface, multi-objective optimization is performed, and the optimization target is iteratively searched for in the optimization space to obtain the optimization model.

However the optimization effect that can be achieved using the optimization model needs to be studied with finite element software to evaluate the transient dynamics simulation calculations. This will allow one to obtain the most effective optimization solution and its effect. The flow of the specific optimization scheme is shown in Figure 12 below.

3.4.2. Sensitivity Screening

By performing the correlation analysis of the parameters of this machine tool model, 131-dimensional variables were finally determined to be optimized, and five optimization targets were identified: the maximum deformation of the main frame (MD-Z), subframe (MD-F), electric control cabinet (MD-K), base bracket (MD-T), and first-order inherent frequency (M1) of the whole frame.

Due to there being numerous size variables, sensitivity screening was performed to obtain the sizes that have a greater impact on the optimization target. The range of test points was set according to the size variable interval, and the Spearman method was used for sensitivity analysis; the closer the correlation value in the results was to one, the greater the influence it had on the optimization target. The dimensions with a correlation greater than 0.5 were screened out, and finally, 15 parameters with the greatest influence on each optimization target were obtained, as shown in

Table 2. Main parameters of sensitivity screening results.

Size Code	Correlation	Maximum Related Object	Interval Value	Instruction
det D	0.56	M1	0–20 mm	Shaft diameter variation
det H	1.00	M1	0–200 mm	Decreased center of gravity
XS1	0.98	MD-T	0.6–1.0	Rib thickness factor of main frame
K1	0.86	MD-T	30–50 mm	Host frame bezel width
CT	0.77	MD-T	30–50 mm	Thickness of top plate of main frame
AT	1.00	MD-Z	30–50 mm	Thickness of front base plate of main frame
XS4	0.57	M1	0.6–1.0	Subframe rib thickness factor
K2	0.64	MD-F	30–50 mm	Sub-rack bezel thickness
K3	0.68	MD-F	30–50 mm	Sub-frame right frame thickness
FAT	0.79	MD-F	15–40 mm	Subframe rear base plate thickness
FDT	0.64	MD-F	20–40 mm	Thickness of subframe top plate
FCT	0.63	MD-F	20–50 mm	Thickness of right bottom plate of subframe
K7	0.73	MD-K	30–50 mm	Electric control cabinet bezel thickness
CCT	0.77	MD-T	20–50 mm	Electric control cabinet base plate thickness
K5	0.53	MD-T	20–50 mm	Thickness of the bottom bracket bezel

below. The positions of variables in the multi-wire saw machine are illustrated in Figure 13.

The remaining model dimensions are set as additional variables according to the relationship between the parameters so that they follow the changes in the 15 main parameters derived from sensitivity screening for the subsequent structural optimization of the whole multi-wire sawing machine.

The design criteria for rib panels as important structures for reducing structural weight and enhancing auxiliary support are as follows:

$$\eta =\frac{h}{XT}\le 5$$

(2)

where $\eta$ denotes the relative height of the rib panels; h denotes the height of the rib panels, and XT denotes the thickness of the base plate.

The thickness of rib panels depends on the thickness of the base plate and the rib thickness coefficient, and their relationship is as follows:

$$t=XT\times XS$$

(3)

where t denotes the thickness of rib panels, and XS denotes the rib thickness factor.

3.4.3. Response Surface Based Multi-Objective Optimization Design

The response surface model is used to obtain the influence relationship between the optimization objective and each input parameter so that structural optimization can be performed based on this mathematical model. Before constructing the response surface, a DOE test is conducted, and here, the OSF_method is used to design the test points, and 287 test points are customized for calculation. After the calculation is completed, the response surface is then constructed using the kriging method, and the correlation error for judging its convergence is set to 5%. When it cannot converge, refinement Points are automatically inserted during the refinement process, and its maximum number of refinement points is set to five. If convergence is achieved during the refinement process, the refinement will be finished.

Due to there being a sufficient number of test points, the refinement error finally reached 2.0047% after using one refinement point for refinement, which satisfied the convergence accuracy requirement, at which time the response surface construction had been completed.

The multi-objective optimal design was based on the constructed response surface. By defining the objective function, setting the constraints, and determining the relationship between the input parameters, the optimal solution satisfying the requirements was found.

The objective function for multi-objective optimization is shown in the following equation.

$$\{\begin{array}{ll}find& \min (P_2)\\ & \min (P_3)\\ & \min (P_4)\\ & \min (P_5)\\ & \max (P_1)\\ seek& \det \mathrm{D}=20\\ & \det \mathrm{H}=200\end{array}$$

(4)

where P₁ denotes the first-order inherent frequency of the multi-wire sawing machine; P₂ denotes the maximum displacement of the main frame; P₃ denotes the maximum displacement of the subframe; P₄ denotes the maximum displacement of the electric control cabinet; P₅ denotes the maximum displacement of the base bracket; det D denotes the change in the spindle axis diameter;and det H denotes the change in the spindle axis height, which is used to measure the lift of the machining center of gravity.

The MOGA method is used for multi-objective iterative optimization search, the initial sample points are set to 100, the number of samples per iteration is set to 100, the maximum number of iterations is set to 20, the stability percentage for judging its convergence is set to 2%, and five candidate points are set to verify the error of response surface calculation, which can also be used as the optimal solution for the calculation. When the iteration occurs for the 12th time, as shown in Figure 14, the convergence stability percentage is 1.98%, which achieves the convergence accuracy and satisfies the requirements of the optimization search. Figure 15 shows the iterative optimization process in multi-objective optimization.

According to the mechanical design manual [20], the rib thickness factor is generally taken as 0.8, and it is made to be consistent with the original, while the first-order intrinsic frequency value of the machine is as high as possible. An optimal solution is selected from five groups of candidate points, and the values are fitted and rounded regarding the other four groups of candidate points. The final values of the parameters of the optimized solution were obtained, as shown in

Table 3. Values of the main parameters of the optimization scheme.

Size Code	Value	Instruction
det D	20 mm	Shaft diameter variation
det H	200 mm	Decreased center of gravity
XS1	0.80	Rib thickness factor of the main frame
K1	50 mm	Host frame bezel width
CT	38 mm	Thickness of top plate of main frame
AT	48 mm	Thickness of front base plate of main frame
XS4	0.94	Subframe rib thickness factor
K2	38 mm	Sub-rack bezel thickness
K3	38 mm	Sub-frame right frame thickness
FAT	36 mm	Subframe rear base plate thickness
FDT	34 mm	Thickness of subframe top plate
FCT	48 mm	Thickness of right bottom plate of subframe
K7	42 mm	Electric control cabinet bezel thickness
CCT	48 mm	Electric control cabinet base plate thickness
K5	48 mm	Thickness of the bottom bracket bezel

below.

A 3D model of the optimized multi-wire sawing machine can be obtained by substituting the values of the parameters obtained in the optimization scheme into the model, and the specific effect is shown in Figure 16 below.

3.4.4. Optimization Program Effect Evaluation

The optimized machine model is re-imported into the finite element software to obtain dynamics verification calculations to realize the verification evaluation of its optimization effect. Finite element analysis was re-executed for the optimized model, and the calculated results are shown in Figure 17 below.

Numerically, the first-order intrinsic frequency of the multi-wire sawing machine in the optimized solution increased from the original 63.62 Hz to 84.15 Hz, an increase of nearly 21 Hz, with a percentage increase of 32.27%, which can reduce the risk of resonance of the machine to a certain extent.

From the data in Figure 17, it can be seen that the optimized excitation frequency at the same line speed is reduced compared with that before optimization. The first-order inherent frequency of the whole machine before optimization is 63.62 Hz, and if running a line speed of 2400 m/min is used, the maximum excitation frequency is 72 Hz. During the acceleration of the multi-wire sawing machine, the excitation frequency will pass through the first-order inherent frequency, and there is a high possibility of resonance.

After optimization, the first-order intrinsic frequency of the machine is 84.15 Hz, and the maximum excitation frequency is 64 Hz at a linear speed of 2400 m/min. The excitation frequency will not pass its first-order intrinsic frequency, so the probability of resonance is low and the stiffness performance of the machine is improved.

Kinetic simulation validation calculations are obtained from the optimized scheme model, and the summary processing of the calculation results is shown in Figure 18 below.

From Figure 6 and Figure 10, it can be seen that the amplitude of each part of the original machine takes the line speed of 1200 m/min as the inflection point, after which the slope increases sharply, while from Figure 18, it can be seen that the slope of the amplitude of each part of the optimized machine increases sharply after the line speed of 2400 m/min, and the inflection point of the amplitude occurs significantly later than that of the pre-optimized machine. Therefore, in terms of the anti-vibration performance, the optimized machine is better than the pre-optimized machine.

The optimization effect of the multi-wire sawing machine in terms of the amplitude of each part is evaluated using specific values, as shown in

Table 4. Comparison of machine amplitudes.

Optimization Objectives		2400 (m/min)		1500 (m/min)
	Original Machine	Optimized Machine	Optimization Percentage	Original Machine
MD-Z (µm)	214.16	61.95	−71.07%	63.26
MD-F (µm)	241.61	56.50	−76.62%	108.36
MD-K (µm)	248.56	82.96	−66.62%	64.23
MD-T (µm)	62.55	14.52	−76.79%	16.28
M1 (Hz)	63.62	84.15	+32.27%	63.62

below.

Table 5. Machine optimization effect equivalence.

Optimization Objectives	Original Machine	Optimized Machine	Optimization Percentage	Conversion Ratio
Wire speed (m/min)	1500	2400	+60%	1.94
Weight (t)	7.09	9.31	+31%

shows the equivalent evaluation of the optimization effect of the optimized solution in terms of line speed and quality.

Froma numerical point of view, the amplitude of each part of the optimized machine is at least 66.62% lower when it runs at a line speed of 2400 m/min, and the amplitude of each part is similar to that of the original machine when it runs at a line speed of 1500 m/min, but the amplitude value of the optimized subframe is much smaller than that of the original machine, and the first-order intrinsic frequency is increased by nearly 21 Hz, which is an increase of 32.27%. The operating stability of the machine has been significantly improved, and the overall performance is better.

4. Conclusions

Although the maximum operating speed of the original machine tool is 1800 m/min, in practice, due to the impact of machine vibration on the quality of slicing, the site is often used at a line speed of 1200 m/min for slicing, and occasionally, a line speed of 1500 m/min is used for some materials that require a lower surface quality. Based on the comparison of the above values, it can be concluded that the machine can be optimized to run a high line speed of 2400 m/min for material sawing.

In terms of results, the optimized machine line speed increased by 60%, and the quality increased by 31%. An increase in quality means an increase inproduction cost, while an increase in line speed means an increase inproduction efficiency and higher economic efficiency at the same time. Therefore, for both comprehensive considerations, the conversion ratio of this optimization reaches 1.94, with a certain quality cost in exchange for a greater increase in sawing efficiency has some significance, which can guide the development of high-line speed multi-wire sawing machines.

• This article first analyzes the vibration principle of a multi-wire sawing machine, exploring the relationship between dynamic stiffness, amplitude, and excitation force. The conclusion is drawn that, under the condition of constant excitation force, improving the dynamic stiffness of the machine tool can effectively reduce the amplitude. In this paper, the transient dynamic finite element simulation calculation of a multi-wire sawing machine is carried out, and the relative error is less than 11.46%, which was compared with the measured data. This can be used to accurately describe the amplitude variation rule with linear velocity and verify the effectiveness and accuracy of the simulation, which is significant in the study of the vibration characteristics of a multi-wire sawing machine under different linear velocities.
• To solve the problem of the large amplitude of a multi-wire sawing machine running at a high wire speed, the structure was optimized using a multi-objective optimization design method based on the response surface, and the optimized results were then validated through dynamic analysis. The final effect was remarkable, and the linear speed was improved by about 60% after optimization. This optimization successfully meets the high-speed cutting requirements of the multi-wire sawing machine.
• Modal analysis was performed on the multi-wire cutting machine to obtain the natural frequencies from the 1st to the 6th order. The reasons for the vibration generated at high line speeds were analyzed. According to the sensitivity screening of the machine, size that has a large influence on the vibration of the machine, and the influence relationship between each parameter variable and the optimization objective is obtained by constructing the response surface. The optimization objectives were set to improve the first natural frequency and reduce the displacement of the frame. Then, research on the vibration control strategy of the machine tool can be carried out accordingly.
• The structure design of a multi-wire sawing machine is summarized: an increase in the diameter of the groove wheel can reduce the motor speed, thus reducing the vibration frequency, reducing the amplitude of each part of the machine tool, reducing the risk of resonance, improving the operation stability of the machine tool, reducing the center of gravity of machine tool, and making its lower end more stable to improve the running stability of the machine tool. Regarding the optimization of the frame structure, the number of ribs, its size, and its position influence its performance.