Construction method, collision simulation method, and system for simplified rail vehicle models
The disclosure provides a construction method including steps of: constructing a detailed model and a simplified model of the rail vehicle using finite element method based on the three-dimensional structure of the rail vehicle; initially setting the simplified model in accordance with the detailed model, including specifying materials, mechanical properties parameters, and mass point; setting different simplification thickness ratio to obtain a plurality of sets of the simplified models, designing longitudinal and lateral collision conditions, and in each type of the collision conditions, performing collision simulations on the detailed model and each set of the simplified models under each thickness ratio to obtain respective collision response curves; calculating a total CORA value corresponding to the plurality of sets of the simplified models; selecting an optimal simplified model from n sets of the simplified models with the highest total CORA values.
1 . A method for constructing a simplified rail vehicle model, comprising:
constructing a detailed model and a simplified model of the rail vehicle using finite element method based on a physical three-dimensional structure of a real rail vehicle, the detailed model of the rail vehicle being a 1:1 scale reproduction vehicle model that replicates the real vehicle's structural details; and the simplified model of the rail vehicle retaining an external form of the real rail vehicle; wherein in the simplified model, a middle part of carriage body of the rail vehicle is configured as a solid single-shell structure, while two ends of the carriage body are configured as hollow double-shell structures;
initializing the simplified model in accordance with the detailed model, including specifying materials, mechanical properties parameters, and mass points;
in response to different simplification thickness ratios input by an input device, generating a plurality of sets of simplified models, each of which proportionally scaling the shell thickness of distinct vehicle body sections based on the corresponding simplification thickness ratio, such that each simplification thickness corresponds each simplified model; designing longitudinal and lateral collision conditions, and in each type of the collision conditions, performing collision simulations on the detailed model and each set of the simplified models under each thickness ratio to obtain respective collision response curves, wherein the thickness ratio is a ratio of the thickness of the solid single-shell of the simplified model to a thickness spacing of the hollow double-shell of the detailed model at a same position;
based on a CORA evaluation method to calculate a total CORA value corresponding to the plurality of sets of the simplified models that have different simplification thickness ratio, based on the collision response curves of the simplified model and the detailed model in a same collision condition, wherein based on a CORA evaluation method to calculate the total CORA value including: calculating CORA values for all the simplified model under the same collision condition based on the collision response curves of the detailed model and the plurality of sets of simplified models under the same collision condition that a dummy model can be directly placed into all the simplified models to conduct secondary collision simulation of train occupants; and performing weighted calculations on the CORA values under all collision conditions to obtain the total CORA value; when there are more than two collision response curves, the CORA value obtained for the simplified model under any of the simplification thickness ratio and for any of the collision conditions is a weighted average of the CORA values obtained for the simplified model under each type of the collision response curves for the corresponding collision condition;
based on the total CORA value, selecting an optimal simplified model from n sets of the simplified models with highest total CORA values, where a value range of n is 2 to 5.
2 . The method for constructing the simplified rail vehicle model according to claim 1 , wherein for the simplified model, when there are common nodes, mesh elements positioned between both ends of the hollow double-shell structure and a central section of the solid single-shell structure are connected through the common nodes; when there are no common nodes, free ends of the hollow double-shell structure are connected to the solid single-shell structure through rigid beam elements.
3 . The method for constructing the simplified rail vehicle model according to claim 1 , wherein the collision response curves further include one or more of a collision kinetic energy-time curve, an internal energy-time curve, and an acceleration-time curve.
4 . The method for constructing the simplified rail vehicle model according to claim 1 , wherein for one type of collision response curve, the CORA value for the simplified model, under any of the simplification thickness ratio in any of the collision conditions, is calculated according to a following formula:
C
=
w
1
C
1
+
w
2
*
C
2
C
2
=
w
2
a
*
C
2
a
+
w
2
b
*
C
2
b
+
w
2
c
*
C
2
c
in the formula, C indicates the CORA value; w 1 , w 2 , w 2a , w 2b , and w 2c are all weighting factors, where the sum of w 1 and w 2 equals 1, and the sum of w 2a , w 2b , and w 2c also equals 1; C 1 indicates a channel correlation score, C 2 indicates a cross-correlation score, C 2a indicates a phase difference score, C 2b indicates a peak difference score, and C 2c indicates a slope difference score.
5 . The method for constructing the simplified rail vehicle model according to claim 1 , wherein operations of selecting the optimal simplified model comprises:
based on the total CORA value of the simplified models with different simplification thickness ratio, selecting the simplified model with the highest similarity in collision deformation mode to the detailed model from the n simplified models with the highest total CORA values as the optimal simplified model.
6 . The method for constructing the simplified rail vehicle model according to claim 4 , wherein operations of obtaining the channel correlation score C 1 comprises:
generating an inner channel and an outer channel based on the collision response curves of the detailed model;
calculating the channel correlation score C 1 based on a positional relationship between the collision response curves of the simplified model and the two signal channels;
wherein when the collision response curve of the simplified model is within a boundary of the inner channel, the channel correlation score C 1 is 1;
when the collision response curve of the simplified model is outside an outer boundary of the outer channel, the channel correlation score C 1 is 0;
when the collision response curve of the simplified model is between the inner and outer channels, the channel correlation score C 1 is calculated according to the following formula:
C
i
=
{
1
,
when
❘
"\[LeftBracketingBar]"
y
(
t
1
)
-
x
(
t
i
)
❘
"\[LeftBracketingBar]"
<
δ
i
(
t
)
(
δ
0
(
t
)
-
❘
"\[LeftBracketingBar]"
y
(
t
i
)
-
x
(
t
i
)
❘
"\[RightBracketingBar]"
δ
0
(
t
)
-
δ
i
(
t
)
)
k
0
,
when
❘
"\[LeftBracketingBar]"
y
(
t
i
)
-
x
(
t
i
)
❘
"\[RightBracketingBar]"
>
δ
i
(
t
)
wherein, k>0 and k∈N
C
1
=
∑
i
=
1
n
C
i
n
,
0
≤
C
1
≤
1
wherein, δ 0 (t) indicates the boundary of the outer channel, δ i (t) indicates the boundary of the inner channel, y(t i ) indicates a reference curve, which is the collision response curve of the detailed model; x(t i ) indicates an evaluation curve, which is the collision response curve of the simplified model C i indicates the score at different time steps, where i indicates the ith time step, k is the exponential factor for the channel correlation score, and N indicates a real number.
7 . The method for constructing the simplified rail vehicle model according to claim 6 , wherein operations of obtaining the cross-correlation score C 2 comprise:
within an evaluation interval [t min , t max ], shifting the reference curve by changing multiples of the time step Δt, and calculating a cross-correlation value between the reference curve and the evaluation curve for each changed time step Δt; wherein the reference curve is the collision response curve of the detailed model, and the evaluation curve is the collision response curve of the simplified model;
at a time step of maximum cross-correlation, calculating the phase difference score C 2a , the slope difference score C 2c , and the peak difference score C 2b between the reference curve and the evaluation curve.
8 . A method for simulating occupant secondary collisions based on a simplified railway vehicle model, comprising:
constructing a detailed model and a simplified model of the rail vehicle using finite element method based on a physical three-dimensional structure of a real rail vehicle, the detailed model of the rail vehicle being a 1:1 scale reproduction vehicle model that replicates the real vehicle's structural details; and the simplified model of the rail vehicle retaining an external form of the real rail vehicle; wherein in the simplified model, a middle part of carriage body of the rail vehicle is configured as a solid single-shell structure, while two ends of the carriage body are configured as hollow double-shell structures,
initializing the simplified model in accordance with the detailed model, including specifying materials, mechanical properties parameters, and mass points;
in response to different simplification thickness ratios input by an input device generating a plurality of sets of simplified models, each of which proportionally scaling the shell thickness of distinct vehicle body sections based on the corresponding simplification thickness ratio, such that each simplification thickness corresponds each simplified model; designing longitudinal and lateral collision conditions, and in each type of the collision conditions, performing collision simulations on the detailed model and each set of the simplified models under each thickness ratio to obtain respective collision response curves, wherein the thickness ratio is a ratio of the thickness of the solid single-shell of the simplified model to a thickness spacing of the hollow double-shell of the detailed model at a same position;
based on a CORA evaluation method to calculate a total CORA value corresponding to the plurality of sets of the simplified models that have different simplification thickness ratio, based on the collision response curves of the simplified model and the detailed model in a same collision condition, wherein based on a CORA evaluation method to calculate the total CORA value including: calculating CORA values for all the simplified model under the same collision condition based on the collision response curves of the detailed model and the plurality of sets of simplified models under the same collision condition that a dummy model can be directly placed into all the simplified models to conduct secondary collision simulation of train occupants; and performing weighted calculations on the CORA values under all collision conditions to obtain the total CORA value; when there are more than two collision response curves, the CORA value obtained for the simplified model under any of the simplification thickness ratio and for any of the collision conditions is a weighted average of the CORA values obtained for the simplified model under each type of the collision response curves for the corresponding collision condition;
based on the total CORA value, selecting an optimal simplified model from n sets of the simplified models with highest total CORA values, where a value range of n is 2 to 5;
using the simplified model of the target railway vehicle to simulate the actual interior decoration and occupant scenarios within the target railway vehicle, and performing collision simulations to obtain collision simulation results.
9 . The method for simulating occupant secondary collisions according to claim 8 , wherein for the simplified model, when there are common nodes, mesh elements positioned between both ends of the hollow double-shell structure and a central section of the solid single-shell structure are connected through the common nodes; when there are no common nodes, free ends of the hollow double-shell structure are connected to the solid single-shell structure through rigid beam elements.
10 . The method for simulating occupant secondary collisions according to claim 8 , wherein for one type of collision response curve, the CORA value for the simplified model, under any of the simplification thickness ratio in any of the collision conditions, is calculated according to the following:
C
=
w
1
C
1
+
w
2
*
C
2
C
2
=
w
2
a
*
C
2
a
+
w
2
b
*
C
2
b
+
w
2
c
*
C
2
c
In the formula, C indicates the CORA value; w 1 , w 2 , w 2a , w 2b , and w 2c are all weighting factors, where the sum of w 1 and w 2 equals 1, and the sum of w 2a , w 2b , and w 2c also equals 1; C 1 indicates a channel correlation score, C 2 indicates a cross-correlation score, C 2a indicates a phase difference score, C 2b indicates a peak difference score, and C 2c indicates a slope difference score.
11 . The method for simulating occupant secondary collisions according to claim 8 , wherein operations of selecting the optimal simplified model comprises:
based on the total CORA value of the simplified models with different simplification thickness ratio, selecting the simplified model with the highest similarity in collision deformation mode to the detailed model from the n simplified models with the highest total CORA values as the optimal simplified model.
12 . The method for simulating occupant secondary collisions according to claim 9 , wherein the collision response curves further include one or more of a collision kinetic energy-time curve, an internal energy-time curve, and an acceleration-time curve.
13 . The method for simulating occupant secondary collisions according to claim 10 , wherein operations of obtaining the channel correlation score C 1 comprises:
generating an inner channel and an outer channel based on the collision response curves of the detailed model;
calculating the channel correlation score C 1 based on a positional relationship between the collision response curves of the simplified model and the two signal channels;
wherein when the collision response curve of the simplified model is within a boundary of the inner channel, the channel correlation score C 1 is 1;
when the collision response curve of the simplified model is outside an outer boundary of the outer channel, the channel correlation score C 1 is 0;
when the collision response curve of the simplified model is between the inner and outer channels, the channel correlation score C 1 is calculated according to the following formula:
C
i
=
{
1
,
when
❘
"\[LeftBracketingBar]"
y
(
t
1
)
-
x
(
t
i
)
❘
"\[LeftBracketingBar]"
<
δ
i
(
t
)
(
δ
0
(
t
)
-
❘
"\[LeftBracketingBar]"
y
(
t
i
)
-
x
(
t
i
)
❘
"\[RightBracketingBar]"
δ
0
(
t
)
-
δ
i
(
t
)
)
k
0
,
when
❘
"\[LeftBracketingBar]"
y
(
t
i
)
-
x
(
t
i
)
❘
"\[RightBracketingBar]"
>
δ
i
(
t
)
wherein, k>0 and k∈N
C
1
=
∑
i
=
1
n
C
i
n
,
0
≤
C
1
≤
1
wherein, δ 0 (t) indicates the boundary of the outer channel, δ i (t) indicates the boundary of the inner channel, y(t i ) indicates a reference curve, which is the collision response curve of the detailed model; x(t i ) indicates an evaluation curve, which is the collision response curve of the simplified model; C i indicates the score at different time steps, where i indicates the ith time step, k is the exponential factor for the channel correlation score, and N indicates a real number.
14 . The method for simulating occupant secondary collisions according to claim 13 , wherein operations of obtaining the cross-correlation score C 2 comprise:
within an evaluation interval [t min , t max ], shifting the reference curve by changing multiples of the time step Δt, and calculating a cross-correlation value between the reference curve and the evaluation curve for each changed time step Δt; wherein the reference curve is the collision response curve of the detailed model, and the evaluation curve is the collision response curve of the simplified model;
at a time step of maximum cross-correlation, calculating the phase difference score C 2a , the slope difference score C 2c , and the peak difference score C 2b between the reference curve and the evaluation curve.
15 . A terminal device, comprising:
a memory, configured to store program instructions; and
a processor, configured to performed the program instructions to perform a method for constructing a simplified rail vehicle model, the method for constructing the simplified rail vehicle model, comprising:
constructing a detailed model and a simplified model of the rail vehicle using finite element method based on a physical three-dimensional structure of a real rail vehicle, the detailed model of the rail vehicle being a 1:1 scale reproduction vehicle model that replicates the real vehicle's structural details; and the simplified model of the rail vehicle retaining an external form of the real rail vehicle; wherein in the simplified model, a middle part of carriage body of the rail vehicle is configured as a solid single-shell structure, while two ends of the carriage body are configured as hollow double-shell structures,
initializing the simplified model in accordance with the detailed model, including specifying materials, mechanical properties parameters, and mass points;
in response to different simplification thickness ratios input by an input device generating a plurality of sets of simplified models, each of which proportionally scaling the shell thickness of distinct vehicle body sections based on the corresponding simplification thickness ratio, such that each simplification thickness corresponds each simplified model; designing longitudinal and lateral collision conditions, and in each type of the collision conditions, performing collision simulations on the detailed model and each set of the simplified models under each thickness ratio to obtain respective collision response curves, wherein the thickness ratio is a ratio of the thickness of the solid single-shell of the simplified model to a thickness spacing of the hollow double-shell of the detailed model at a same position;
based on a CORA evaluation method to calculate a total CORA value corresponding to the plurality of sets of the simplified models that have different simplification thickness ratio, based on the collision response curves of the simplified model and the detailed model in a same collision condition, wherein based on a CORA evaluation method to calculate the total CORA value including: calculating CORA values for all the simplified model under the same collision condition based on the collision response curves of the detailed model and the plurality of sets of simplified models under the same collision condition that a dummy model can be directly placed into all the simplified models to conduct secondary collision simulation of train occupants; and performing weighted calculations on the CORA values under all collision conditions to obtain the total CORA value; when there are more than two collision response curves, the CORA value obtained for the simplified model under any of the simplification thickness ratio and for any of the collision conditions is a weighted average of the CORA values obtained for the simplified model under each type of the collision response curves for the corresponding collision condition;
based on the total CORA value, selecting an optimal simplified model from n sets of the simplified models with highest total CORA values, where a value range of n is 2 to 5.
16 . The terminal according to claim 15 , wherein for the simplified model, when there are common nodes, mesh elements positioned between both ends of the hollow double-shell structure and a central section of the solid single-shell structure are connected through the common nodes; when there are no common nodes, free ends of the hollow double-shell structure are connected to the solid single-shell structure through rigid beam elements.
17 . The terminal according to claim 15 , wherein for one type of collision response curve, the CORA value for the simplified model, under any of the simplification thickness ratio in any of the collision conditions, is calculated according to the following:
C
=
w
1
C
1
+
w
2
*
C
2
C
2
=
w
2
a
*
C
2
a
+
w
2
b
*
C
2
b
+
w
2
c
*
C
2
c
in the formula, C indicates the CORA value; w 1 , w 2 , w 2a , w 2b , and w 2c are all weighting factors, where the sum of w 1 and w 2 equals 1, and the sum of w 2a , w 2b , and w 2c also equals 1; C 1 indicates a channel correlation score, C 2 indicates a cross-correlation score, C 2a indicates a phase difference score, C 2b indicates a peak difference score, and C 2c indicates a slope difference score.
18 . The terminal according to claim 16 , wherein the collision response curves further include one or more of a collision kinetic energy-time curve, an internal energy-time curve, and an acceleration-time curve.
19 . The terminal according to claim 17 , wherein operations of obtaining the channel correlation score C 1 comprise:
generating an inner channel and an outer channel based on the collision response curves of the detailed model;
calculating the channel correlation score C 1 based on a positional relationship between the collision response curves of the simplified model and the two signal channels;
wherein when the collision response curve of the simplified model is within a boundary of the inner channel, the channel correlation score C 1 is 1;
when the collision response curve of the simplified model is outside an outer boundary of the outer channel, the channel correlation score C 1 is 0;
when the collision response curve of the simplified model is between the inner and outer channels, the channel correlation score C 1 is calculated according to the following formula:
C
i
=
{
1
,
when
❘
"\[LeftBracketingBar]"
y
(
t
1
)
-
x
(
t
i
)
❘
"\[LeftBracketingBar]"
<
δ
i
(
t
)
(
δ
0
(
t
)
-
❘
"\[LeftBracketingBar]"
y
(
t
i
)
-
x
(
t
i
)
❘
"\[RightBracketingBar]"
δ
0
(
t
)
-
δ
i
(
t
)
)
k
0
,
when
❘
"\[LeftBracketingBar]"
y
(
t
i
)
-
x
(
t
i
)
❘
"\[RightBracketingBar]"
>
δ
i
(
t
)
Wherein, k>0 and k∈N
C
1
=
∑
i
=
1
n
C
i
n
,
0
≤
C
1
≤
1
Wherein, δ 0 (t) indicates the boundary of the outer channel, δ i (t) indicates the boundary of the inner channel, y(t i ) indicates a reference curve, which is the collision response curve of the detailed model; x(t i ) indicates an evaluation curve, which is the collision response curve of the simplified model; C i indicates the score at different time steps, where i indicates the ith time step, k is the exponential factor for the channel correlation score, and N indicates a real number.
20 . The terminal according to claim 19 , wherein operations of obtaining the cross-correlation score C 2 comprises:
within an evaluation interval [t min , t max ], shifting the reference curve by changing multiples of the time step Δt, and calculating a cross-correlation value between the reference curve and the evaluation curve for each changed time step Δt; wherein the reference curve is the collision response curve of the detailed model, and the evaluation curve is the collision response curve of the simplified model;
at a time step of maximum cross-correlation, calculating the phase difference score C 2a , the slope difference score C 2c , and the peak difference score C 2b between the reference curve and the evaluation curve.