Created on:2021-07-08 14:14

Hydraulic fault diagnosis and maintenance of rubber briquetting machine

Hydraulic fault diagnosis and maintenance of rubber briquetting machine

(1) Hydraulic system of rubber briquetting machine

According to the working requirements of the rubber post-treatment production line, the hydraulic system of the rubber briquetting machine adopts two-way cartridge valve circuit, the components adopt pressure seal, and the action test is completed on the processing site. The hydraulic system is shown in Figure o. The two-way cartridge valve circuit can realize the differential quick action of the hydraulic cylinder. It can adjust the function according to the actual production requirements, and also can easily adjust the action sequence of each hydraulic cylinder.

(2) Fault diagnosis and search of pump motor unit

① The fault tree analysis of pump motor group stop in the process of reliability growth of the hydraulic system of rubber briquetting machine, it is found that the pump noise is large and abnormal, and the system stops suddenly. Fault tree analysis is used for fault diagnosis. Because each component is relatively independent in the system, each component can be regarded as a fault subsystem to determine the intermediate events and build the tree step by step. According to this idea, a fault tree can be established with the top event of the pump motor group stop as shown in Figure P.

Fault tree analysis of the system, and step by step down the analysis, inspection, found that there is no problem with the electrical system. The problem turned to the middle event of motor overload. After analyzing and checking the pump circuit and removing the large pump coupling, the small pump and the system worked normally. The focus of the problem focused on the large pump circuit. By analyzing downward in turn, the spool of the two-way cartridge valve 8 corresponding to the large pump was removed. After starting, the pump motor unit operated normally, indicating that the large pump itself had no fault. According to the experience and the lethality of the components, the failure rate of the valve components is high, so the priority is to check the relief valve, and it is found that the control damping in the relief valve of the large pump is blocked, which leads to the abnormal operation of the large pump and makes the pressure of the large pump out of control. This is not conducive to the safety of large pumps and systems. However, the protection measures such as motor overload and pressure overshoot of relief valve are adopted in the design of the system to reduce the harm of the fault.


The direct reason for the blockage of the overflow valve of the large pump is the pollution of the working medium. The pollutants are stuck in the overflow valve, which leads to the abnormal opening of the overflow valve. There is a control damping hole of 0.3mm in the overflow valve, while the control damping hole of 0.5mm in the overflow valve of small pump, which indicates that the damping hole is blocked or blocked by oil pollution. Through oil sample inspection, it is found that the oil pollution is serious, which indicates that this judgment is correct. Check the filter and find that the filter element is invalid. This is because the flow rate in the oil return pipeline is large and the flow rate is too high, which leads to the high back pressure and the failure of the filter element. The above faults can be solved by flushing the filter and replacing the large flow filter element.

② The normalized search decision matrix assumes that there are n search schemes and M attributes that affect the search scheme, such as failure probability and search cost. X = (x1, X2,..., xn) is used to represent the set of alternative schemes, and Yi = (Yi1, yi2,..., Yim) is used to represent the set of attribute values of the ith scheme, where Yij is the value of the j-th attribute of the ith scheme. Then the attribute Yij = FJ (XI), I = 1,2 ·, N; j=1,2,…,m。 The attribute values of each search scheme can be represented by search decision matrix A, which provides basic information for analyzing decision problems.

A= (A)

According to the fault tree shown in Figure P, the decision matrix is determined. Xi stands for fault reason and number. Only the attributes that affect the decision of search strategy are search cost C and failure probability P. the order is shown in table a. In the table, if the occurrence probability p is 1 (X7 and x9 fault causes), the occurrence probability is the highest, and if the search cost C is 1, the search cost is the lowest, and so on.

According to table a, the search decision matrix is

A= (B)


Search cost and failure probability ranking of failure causes


According to the normalization theory, the attributes are normalized, that is, the values of each attribute are uniformly transformed into the range of (0,1), which becomes the normalized matrix Z, in which the element Z is

Zij=yij i=1,2,···,n; j=1,2,···,m (C)

Normalize the matrix according to formula (c)


③ Weighted normalized search decision matrix this method uses the least square method of weight to determine the weight of attributes, and compares the importance of each attribute in pairs. If the estimated value of the importance of the i-th attribute relative to the j-th attribute is recorded as bij, and it is considered that it is approximately the ratio wi / WJ of the weight of attribute I and the weight of attribute I, then the result of pairwise comparison of N attributes is represented by matrix B, and the result is obtained

B= (2-5)

If the estimates of bij (I, j = 1, 2,..., m) are consistent, then

bij=1/bij bij=1/bikbkj (2-6)

If the estimates of bij are inconsistent, then

bij≈Wi/Wj (2-7)

In general, the value of bijwj wi is not zero, but a set of weights {W1, W2,..., WM} can be selected to minimize the sum of the squares of errors

min (2-8)

The weight {W1, W2,..., WM) in equation (2-8) is constrained by


Then the Lagrange function is

L= (2-10)

The differential of the above formula to W is obtained


Equation (2-9) and equation (2-11) constitute m + 1 equations and M + 1 unknowns( λ, W1, W2,..., WM). The weighted norm search decision matrix X can be obtained

Xij =WjZij i=1,2,…,n; j=l,2,…,m

Combined with the characteristics of hydraulic system of briquetting machine, according to the actual fault diagnosis, the system compares the weight of search cost and fault probability attribute in pairs to get the matrix


Get the weighted matrix

W= [0.333,0.667]T

The weighted normalized search decision matrix is obtained


④ The search scheme is determined by the sorting method which is close to the ideal solution, and the ideal solution and negative ideal solution are determined. The ideal solution is the best solution x, that is, the optimal search scheme, whose attribute values reach the best value among the candidate schemes; The negative ideal solution is the worst solution x, and its attribute value reaches the worst value among the candidate schemes.

x*={(max Xij)|j∈J|i=1,2,···,n} (2-13)

X- ={ (min Xij) |j∈Jli=l,2,…,n} (2-14)

The distances from each solution to the ideal solution and the negative ideal solution are






The relative closeness of each solution to the ideal solution is

Ei* =Di-/(Di-+Di*) O≤Ei*≤1 (2-16)

The priority of the search scheme is in the order of EI * from large to small, and the first search scheme should give priority to search and detect. First, the ideal solution and negative ideal solution are determined

X*={0. 116,0.077}

X-={0. 019,0.232}

The distances from each solution to ideal solution and negative ideal solution are

Di*={0. 164,0.208,…,0.042}


The relative closeness of each solution to the ideal solution is

Ei*={0.212,0,0.449,0.449, 0.587,0.587,1,0.813,1,0.813}

According to EI * from large to small, the search scheme should be determined, and the first scheme should be given priority to search and diagnosis. According to the above calculation, the order of search schemes is X7, x9, X8, X10, X5, X6, X3, x4, x1, x2.

Home    Article    Hydraulic fault diagnosis and maintenance of rubber briquetting machine