A mathematical model-based method for predicting roll wear in bar rolling.
By establishing a roll wear prediction method based on mathematical model, considering factors such as rolling force, roll diameter and time, the problem of inaccurate roll wear prediction in the existing technology is solved, and the dimensional control accuracy and production safety of rod rolling are improved. sex.
Patent Information
- Application Number
- CN202210084123.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-01-25
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2042-01-25
AI Technical Summary
The prior art has inaccurate roll wear prediction in rod rolling, which makes it difficult to control dimensionally and affects product quality and production safety. The main reason is that factors such as rolling force, steel grade and roller diameter cannot be effectively considered.
Establish a roll wear prediction method based on mathematical model. By calculating the rolling force per unit width, the relative radius of the rolling roll and the rolling time, combined with the computer iterative cycle search weight factor, a wear model for online prediction is constructed.
It improves the accuracy of roll wear prediction, provides a basis for manual estimation and automatic control, reduces the phenomenon of steel pulling, and improves the accuracy of dimensional control.
Smart Images

Figure CN114492024B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of bar rolling, relates to a method for predicting roller wear in bar rolling, and specifically relates to a method for predicting roller wear in bar rolling based on a mathematical model. Background Art
[0002] Bar rolling is a multi-stand continuous rolling groove rolling, which has the characteristics of both continuous rolling and groove rolling. Therefore, if the top and bottom dimensions of a stand are not accurately controlled, it will directly lead to the mismatch between the second flow of the stand and other stands, which will lead to the phenomenon of steel pile-up, affecting product quality and even production safety.
[0003] The top and bottom dimensions of bar rolling are mainly affected by several factors, including the standard depth of the groove bottom, preset roll gap, bounce, and roll wear. The standard depth of the groove bottom can be directly obtained from the groove design drawing, the preset roll gap can be set manually, and the bounce value can be directly calculated by dividing the rolling force model by the rolling mill stiffness (and the bounce of most rolling mills is generally less than 0.5mm, which is relatively stable, with little fluctuation, and the rules are relatively easy to grasp).
[0004] However, for roller wear, most front-line workers in most companies currently estimate the wear value based only on the rolling tonnage, which often results in a large difference between the estimated value and the measured value. The main reasons are as follows: First, different rack positions have different steel temperatures and different reductions, which lead to different rolling forces and therefore different wear; second, wear is also greatly affected by the type of steel. Even for different types of steel at the same rack position, due to the large difference in deformation resistance, the rolling force deviation can reach more than 60%, so their wear conditions are also quite different; third, roller wear is greatly affected by the roller diameter. New rollers are more wear-resistant due to the presence of a hardened layer, while for old rollers, as the depth of the hardened layer increases, its hardness decreases, so its wear resistance is poor.
[0005] Based on the above background, it is necessary to develop a more accurate roll wear prediction method to improve the dimensional control accuracy of bar rolling and reduce the phenomenon of steel piling. Summary of the invention
[0006] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a method for predicting roller wear for bar rolling based on a mathematical model. The method establishes a mathematical model for online prediction of roller wear under the premise of comprehensively considering the factors affecting roller wear. Based on the mathematical model, it can provide a reference for manual estimation of wear values and provide basic support for automatic control of the size of bar rolling.
[0007] The present invention is achieved through the following technical solutions:
[0008] The method for predicting roll wear of bar rolling based on mathematical model includes the following steps: considering the rolling force per unit width , relative radius of the roller And rolling time Under the condition of comprehensive influence on roll wear, a mathematical model for online prediction of roll wear is established, as shown in the following formula:
[0009] ;
[0010] Detect or calculate the relevant parameters of the test frame, and on this basis, calculate the rolling force per unit width of the test frame, the relative radius of the roller and the rolling time, substitute the obtained data into the above mathematical model, and calculate the roller wear value of the test frame.
[0011] A further improvement of the present invention is:
[0012] Consider the effect of rolling force per unit width on roll wear: It is generally believed that the greater the rolling force, the greater the roll wear value. However, if the rolling force is large and the bar diameter is also large, the rolling force distributed to the roll per unit width may not be large. Therefore, the rolling force factor that affects roll wear should be considered as the rolling force per unit width.
[0013] Rolling force per unit width It is the rolling force F divided by the contact width B, where the rolling force is calculated using the Sims formula, specifically:
[0014] ;
[0015] Where, is the deformation resistance of the rolled steel, and its specific value can be calculated through the steel composition; B is the width of the rolled piece, i.e. the contact width; R is the roller radius; is the amount of depression; is the contact arc length;
[0016] It follows that the rolling force per unit width The calculation formula is as follows:
[0017] ,
[0018] Right now:
[0019] .
[0020] Furthermore, the influence of the relative radius of the roll on the wear of the roll is considered: since the new roll is quenched, there is a hardened layer on its surface, and it is easy to understand from the knowledge of heat treatment that the closer the hardened layer is to the outer surface, the higher its martensite content is, and therefore the greater its hardness is. Therefore, the wear resistance of the roll will gradually decrease with the decrease of the radius. The present invention adopts the relative radius of the roll The state of the current roller radius is shown in the following formula:
[0021]
[0022] In the formula, R max is the new roller radius, and R is the radius currently used. The larger it is, the smaller the current roller radius is and the less wear-resistant it is; The smaller it is, the closer the current roll is to a new roll and the more wear-resistant it is.
[0023] Furthermore, the effect of rolling time on roll wear is considered: the longer the pass time t is, the greater the wear is; since bar rolling is characterized by continuous rolling, the flow rate per second of each stand is equal, so for a fixed volume of bar, the time it takes to pass through each stand is equal. Based on this, it can be obtained that the pass time t of each stand is equal to the pass time of the last stand.
[0024] Rolling time That is, the hole feeding time, which can be expressed by the last rack feeding time, as shown in the following formula:
[0025]
[0026] Where V 坯 is the blank volume, S 棒 is the cross-sectional area of the bar final rolling product, and v is the linear speed of the bar final rolling product. It is easy to understand. is the length of the final rolled product of the bar; therefore, It is the feeding time of the last rack.
[0027] In summary, the present invention establishes the following roller wear model, where w is the roller wear prediction value, and the specific model is as follows:
[0028]
[0029] (1)
[0030] In the formula, α0~α3 are weighted influencing factors, which are used to fully represent the possible nonlinear effects of various factors on wear. The appropriate α0~α3 can be searched by computer iterative cycle based on equipment parameters and production parameters combined with the actual wear values measured.
[0031] The specific process of the iterative cycle is as follows: for a specific stand, each time a bar is rolled, the rolling force per unit width is calculated. , relative radius of the roller And rolling time And store, perform n measurements, that is, the actual wear value W corresponding to n roller periods实 , and the true values of its related influencing factors. According to the formula structure shown in Formula 1, the iterative search range of α0~α3 is set and iterative calculation is performed. It is easy to understand that for each given set of α0, α1, α2, α3, combined with Formula 1 and the stored rolling force per unit width , relative radius of the roller And rolling time , a calculated value of roller wear W can be calculated 计 .
[0032] Among them, the search range of α0, α1, α2, and α3 is set as follows:
[0033] Determine the search range of α0: First, according to the order of magnitude of the wear values detected by our factory, such as the average wear value x0, determine the center value of α0. Then, let the search range of α0 be (-3x 0, 3x0).
[0034] Determine the search range of α1, α2, and α3: their central values are all 1, that is, all three of them fluctuate around 1. Therefore, taking α1 as an example, let =1, that is, according to The average value of α1 can be obtained by finding the central value x1 of α1, and the search range of α1 is (-3x 1, 3x1) is sufficient. The method for determining the search range of α2 and α3 is the same as above.
[0035] Based on the above process, in the process of using computer iterative cycle to search for suitable α0~α3, define each group of wear measured values W 实 And the wear calculation value W 计 The distance formula Distance is as follows:
[0036] ;
[0037] Among them, α0~α3 corresponding to the minimum Distance value is the optimal α0~α3 value.
[0038] The roller wear model provided in the present invention is a basic formula, and other influencing factors, such as roller material, etc., can be added according to actual conditions.
[0039] The beneficial effects of the present invention are:
[0040] Under the premise of comprehensively considering the factors affecting roll wear, a mathematical model for online prediction of roll wear was established. Based on this mathematical model, it can provide a reference for manual estimation of wear values and provide basic support for automatic dimensional control of bar rolling.
[0041] The influencing factors of the model of the present invention can be increased or decreased according to the current status of different enterprises, and its optimal parameters α0~α3 can also be modified according to the different equipment and process parameters of the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] Figure 1 Flowchart of iterative loop of the present invention. DETAILED DESCRIPTION
[0043] The method for predicting roll wear of bar rolling based on mathematical model includes the following steps: considering the rolling force per unit width , relative radius of the roller And rolling time Under the condition of comprehensive influence on roll wear, a mathematical model for online prediction of roll wear is established, as shown in the following formula:
[0044] ;
[0045] Detect or calculate the relevant parameters of the test frame, and on this basis, calculate the rolling force per unit width of the test frame, the relative radius of the roller and the rolling time, substitute the obtained data into the above mathematical model, and calculate the roller wear value of the test frame.
[0046] Consider the effect of rolling force per unit width on roll wear: It is generally believed that the greater the rolling force, the greater the roll wear value. However, if the rolling force is large and the bar diameter is also large, the rolling force distributed to the roll per unit width may not be large. Therefore, the rolling force factor that affects roll wear should be considered as the rolling force per unit width.
[0047] Rolling force per unit width It is the rolling force F divided by the contact width B, where the rolling force is calculated using the Sims formula, specifically:
[0048] ;
[0049] Where, is the deformation resistance of the rolled steel, and its specific value can be calculated through the steel composition; B is the width of the rolled piece, i.e. the contact width; R is the roller radius; is the amount of depression; is the contact arc length;
[0050] It follows that the rolling force per unit width The calculation formula is as follows:
[0051] ,
[0052] Right now:
[0053] .
[0054] Consider the influence of the relative radius of the roll on the wear of the roll: Since the new roll is quenched, there is a hardened layer on its surface. It is easy to understand from the knowledge of heat treatment that the closer the hardened layer is to the outer surface, the higher its martensite content is, and therefore the greater its hardness is. Therefore, the wear resistance of the roll will gradually decrease with the decrease of the radius. The present invention adopts the relative radius of the roll. The state of the current roller radius is shown in the following formula:
[0055]
[0056] In the formula, R max is the new roller radius, and R is the radius currently used. The larger it is, the smaller the current roller radius is and the less wear-resistant it is; The smaller it is, the closer the current roll is to a new roll and the more wear-resistant it is.
[0057] Consider the effect of rolling time on roll wear: the longer the pass time t, the greater the wear; because bar rolling has the characteristics of continuous rolling, the flow rate per second of each stand is equal, so for a fixed volume of bar, the time it takes to pass through each stand is equal. Based on this, it can be obtained that the pass time t of each stand is equal to the pass time of the last stand.
[0058] Rolling time That is, the hole feeding time, which can be expressed by the last rack feeding time, as shown in the following formula:
[0059]
[0060] Where V 坯 is the blank volume, S 棒 is the cross-sectional area of the bar final rolling product, and v is the linear speed of the bar final rolling product. It is easy to understand. is the length of the final rolled product of the bar; therefore, It is the feeding time of the last rack.
[0061] In summary, the present invention establishes the following roller wear model, where w is the roller wear prediction value, and the specific model is as follows:
[0062]
[0063] (1)
[0064] In the formula, α0~α3 are weighted influencing factors, which are used to fully represent the possible nonlinear effects of various factors on wear. The appropriate α0~α3 can be searched by computer iterative cycle based on equipment parameters and production parameters combined with the actual wear values measured.
[0065] The specific process of the iterative cycle is as follows: for a specific stand, each time a bar is rolled, the rolling force per unit width is calculated. , relative radius of the roller And rolling time And store, perform n measurements, that is, the actual wear value W corresponding to n roller periods 实 , and the true values of its related influencing factors. According to the formula structure shown in Formula 1, the iterative search range of α0~α3 is set and iterative calculation is performed. It is easy to understand that for each given set of α0, α1, α2, α3, combined with Formula 1 and the stored rolling force per unit width , relative radius of the roller And rolling time , a calculated value of roller wear W can be calculated 计 .
[0066] Among them, the search range of α0, α1, α2, and α3 is set as follows:
[0067] Determine the search range of α0: First, according to the order of magnitude of the wear values detected by our factory, such as the average wear value x0, determine the center value of α0. Then, let the search range of α0 be (-3x 0, 3x0).
[0068] Determine the search range of α1, α2, and α3: their central values are all 1, that is, all three of them fluctuate around 1. Therefore, taking α1 as an example, let =1, that is, according to The average value of α1 can be obtained by finding the central value x1 of α1, and the search range of α1 is (-3x 1, 3x1) is sufficient. The method for determining the search range of α2 and α3 is the same as above.
[0069] Taking the No. 5 stand used by our company as an example, the values corresponding to the n roll periods are shown in Table 1:
[0070]
[0071] Table 1 Storage structure of influencing factors
[0072] Based on the above process, in the process of using computer iterative cycle to search for suitable α0~α3, define each group of wear measured values W 实 And the wear calculation value W 计 The distance formula Distance is as follows:
[0073] ;
[0074] Among them, α0~α3 corresponding to the minimum Distance value is the optimal α0~α3 value.
[0075] In summary, the present invention analyzes the influencing factors of roll wear and their calculation formulas, constructs the distance equation between the true value and the measured value, and combines the computer iterative search method to develop a mathematical model of roll wear. The influencing factors of this model can be easily increased or decreased according to the current status of different enterprises, and its optimal parameters α0~α3 can also be modified according to the different equipment and process parameters of the enterprise.
Claims
1. A method for predicting roll wear in bar rolling based on a mathematical model, characterized in that: The following steps are involved: Considering the rolling force per unit width F B , relative radius R of the roller r Under the conditions of the comprehensive influence of rolling time t on roll wear, a mathematical model for online prediction of roll wear is established, as shown in the following formula: w=α0*(F B ) α1 *(R r ) α2 *t α3 ; Detect or calculate the relevant parameters of the test frame, and on this basis, calculate the rolling force per unit width of the test frame, the relative radius of the roll and the rolling time, substitute the obtained data into the above mathematical model, and calculate the roll wear value of the test frame; where α0~α3 are weighted influencing factors, and according to the equipment parameters and production parameters, combined with the measured actual wear value, use a computer iterative cycle method to search for suitable α0~α3; The rolling force per unit width F B It is the rolling force F divided by the contact width B, where the rolling force is calculated using the Sims formula, specifically: In the formula, σ is the deformation resistance of the rolled steel, and its specific value is calculated based on the steel composition; B is the width of the rolled piece, i.e. the contact width; R is the roller radius; Δh is the reduction; is the contact arc length; It follows that the rolling force per unit width F B The calculation formula is as follows: Right now: The relative radius R of the roller r The state of the current roller radius is shown in the following formula: In the formula, R max is the new roller radius, R is the radius currently used; The rolling time t is the hole feeding time, which is expressed by the last stand feeding time, as shown in the following formula: Where V 坯 is the blank volume, S 棒 is the cross-sectional area of the bar final rolled product, and v is the linear speed of the bar final rolled product. It is easy to understand. is the length of the final rolled product of the bar; therefore, It is the feeding time of the last rack.
2. The method for predicting roll wear of bar rolling based on mathematical model according to claim 1, characterized in that: In the process of searching for suitable α0~α3 by computer iteration cycle, define each group of wear measured values W 实 And the wear calculation value W 计 The distance formula Distance is as follows: Among them, α0~α3 corresponding to the minimum Distance value are the optimal α0~α3 values.
