Analytical reference cases

This chapter belongs to the detailed evidence of the Solver verification. It compares all cases for which the validation catalog contains a closed-form solution with the SpanDraft calculation.

Derivation of the references

The references begin with equilibrium of the overall system. From it follow the support reactions and the moment curve M(x)M(x). The deflection curve arises by integrating the beam equation twice, with the constants of integration from the boundary conditions.1

EI w′′(x)=M(x)EI\,w''(x)=M(x)
(1)

The calculation runs with exact rational numbers and calls SpanDraft nowhere. Only for storing are the results converted to floating-point numbers. The reference is thus independent of the solver but deliberately rests on the same beam theory. It checks the solution of the model and not the model.

The formulas below state the loads FF, qq and CC as magnitudes; their direction is given in the respective caption. All results follow the SpanDraft sign convention: ww is positive upward, and a sagging moment is positive. Cases V01 to V06 use EI=1.68⋅106 Nm2EI = 1.68\cdot 10^6\,\mathrm{Nm^2}. In the figures, moments are given in Nm and deflections in m. The curves are scaled to their largest magnitude in each figure.

What is compared per case

The table at the end of the chapter lists 87 or 63 individual comparisons per case. Case V01 is composed as follows:

ComparisonScopeNumber
Fields12 evaluation points with uu, ww, θ\theta, NN, VV and MM each72
Support reactionsRxR_x and RyR_y at the pinned support, RyR_y at the roller support3
Extreme valuesminimum and maximum of ww, VV and MM6
Extreme locationslocation and side of these six extreme values6

Evaluation points lie at the beam ends, at supports, at load locations and at the boundaries of distributed loads; at interior locations of this kind, left and right are separate. Additional evaluation points are placed at the relative positions x/L=0.137; 0.333; 0.618; 0.873x/L=0.137;\ 0.333;\ 0.618;\ 0.873 and, in each subinterval between two such locations, the points at 37 % and 63 %. The fields are thus also checked between the nodes.

Simply supported and cantilever

V01: central point force

At the load location, the moment has a kink and the shear force a jump.

Simply supported beam with a downward point force at midspan, below it the triangular moment curve with 1200 Nm and the deflection curve with minus 0.952 times 10 to the power of minus 3 meters at the middle.
Figure 1: Case V01 with L = 4 m and F = 1200 N downward.

Analytical

Ra=Rb=F2=600 NR_\mathrm{a}=R_\mathrm{b}=\frac{F}{2}=600\,\mathrm{N}

Mmax=FL4=1200 NmM_\mathrm{max}=\frac{FL}{4}=1200\,\mathrm{Nm}

wmin=−FL348EI=−0.952381 mmw_\mathrm{min}=-\frac{FL^3}{48EI}=-0.952381\,\mathrm{mm} at x=2 mx=2\,\mathrm{m}

SpanDraft

Ra=Rb=600 NR_\mathrm{a}=R_\mathrm{b}=600\,\mathrm{N}

Mmax=1200 NmM_\mathrm{max}=1200\,\mathrm{Nm}

wmin=−0.952381 mmw_\mathrm{min}=-0.952381\,\mathrm{mm} at x=2 mx=2\,\mathrm{m}

V02: cantilever with end force

The fixed support provides a force reaction and a moment reaction.

Cantilever fixed at the left with a point force at the free end, below it the linear moment curve with minus 2400 Nm at the fixed support and the deflection curve with minus 4.286 times 10 to the power of minus 3 meters at the free end.
Figure 2: Case V02 with L = 3 m and F = 800 N downward.

Analytical

Ry=F=800 NR_y=F=800\,\mathrm{N}

RM=FL=2400 NmR_M=FL=2400\,\mathrm{Nm}

Mmin=−FL=−2400 NmM_\mathrm{min}=-FL=-2400\,\mathrm{Nm}

wmin=−FL33EI=−4.28571 mmw_\mathrm{min}=-\frac{FL^3}{3EI}=-4.28571\,\mathrm{mm} at x=3 mx=3\,\mathrm{m}

SpanDraft

Ry=800 NR_y=800\,\mathrm{N}

RM=2400 NmR_M=2400\,\mathrm{Nm}

Mmin=−2400 NmM_\mathrm{min}=-2400\,\mathrm{Nm}

wmin=−4.28571 mmw_\mathrm{min}=-4.28571\,\mathrm{mm} at x=3 mx=3\,\mathrm{m}

V03: cantilever under distributed load

The moment is quadratic, the deflection curve a polynomial of fourth degree.

Cantilever fixed at the left under constant distributed load, below it the parabolic moment curve with minus 2250 Nm at the fixed support and the deflection curve with minus 3.013 times 10 to the power of minus 3 meters at the free end.
Figure 3: Case V03 with L = 3 m and q = 500 N/m downward.

Analytical

Ry=qL=1500 NR_y=qL=1500\,\mathrm{N}

RM=qL22=2250 NmR_M=\frac{qL^2}{2}=2250\,\mathrm{Nm}

Mmin=−qL22=−2250 NmM_\mathrm{min}=-\frac{qL^2}{2}=-2250\,\mathrm{Nm}

wmin=−qL48EI=−3.01339 mmw_\mathrm{min}=-\frac{qL^4}{8EI}=-3.01339\,\mathrm{mm} at x=3 mx=3\,\mathrm{m}

SpanDraft

Ry=1500 NR_y=1500\,\mathrm{N}

RM=2250 NmR_M=2250\,\mathrm{Nm}

Mmin=−2250 NmM_\mathrm{min}=-2250\,\mathrm{Nm}

wmin=−3.01339 mmw_\mathrm{min}=-3.01339\,\mathrm{mm} at x=3 mx=3\,\mathrm{m}

V04: simply supported under distributed load

There is no station at midspan. Both extreme values therefore lie in the interior of a piece.

Simply supported beam under constant distributed load, below it the parabolic moment curve with 1200 Nm and the deflection curve with minus 1.190 times 10 to the power of minus 3 meters at midspan.
Figure 4: Case V04 with L = 4 m and q = 600 N/m downward.

Analytical

Ra=Rb=qL2=1200 NR_\mathrm{a}=R_\mathrm{b}=\frac{qL}{2}=1200\,\mathrm{N}

Mmax=qL28=1200 NmM_\mathrm{max}=\frac{qL^2}{8}=1200\,\mathrm{Nm}

wmin=−5qL4384EI=−1.19048 mmw_\mathrm{min}=-\frac{5qL^4}{384EI}=-1.19048\,\mathrm{mm} at x=2 mx=2\,\mathrm{m}

SpanDraft

Ra=Rb=1200 NR_\mathrm{a}=R_\mathrm{b}=1200\,\mathrm{N}

Mmax=1200 NmM_\mathrm{max}=1200\,\mathrm{Nm}

wmin=−1.19048 mmw_\mathrm{min}=-1.19048\,\mathrm{mm} at x=2 mx=2\,\mathrm{m}

Jumps and extreme values next to the load location

V05: interior point moment

The moment jumps by −C-C at the load location. The deflection curve remains continuous and smooth and lies above the initial position over the entire length.

Simply supported beam with a point moment at 1.5 m, below it the moment curve with a jump from 262.5 to minus 437.5 Nm and the continuous deflection curve with a positive maximum of 0.188 times 10 to the power of minus 3 meters.
Figure 5: Case V05 with L = 4 m and C = 700 Nm counterclockwise at a = 1.5 m.

Analytical

Ra=−Rb=CL=175 NR_\mathrm{a}=-R_\mathrm{b}=\frac{C}{L}=175\,\mathrm{N}

M(a−)=CaL=262.5 NmM(a^-)=\frac{Ca}{L}=262.5\,\mathrm{Nm}

M(a+)=−C(L−a)L=−437.5 NmM(a^+)=-\frac{C(L-a)}{L}=-437.5\,\mathrm{Nm}

wmax=0.187992 mmw_\mathrm{max}=0.187992\,\mathrm{mm} at x=2.24406 mx=2.24406\,\mathrm{m} from θ(x)=0\theta(x)=0

SpanDraft

Ra=−Rb=175 NR_\mathrm{a}=-R_\mathrm{b}=175\,\mathrm{N}

M(a−)=262.5 NmM(a^-)=262.5\,\mathrm{Nm}

M(a+)=−437.5 NmM(a^+)=-437.5\,\mathrm{Nm}

wmax=0.187992 mmw_\mathrm{max}=0.187992\,\mathrm{mm} at x=2.24406 mx=2.24406\,\mathrm{m}

V06: asymmetric point force

The largest sag does not lie under the load, and there is no station at its location.

Simply supported beam with a point force at 1.2 m, below it the kinked moment curve with 1092 Nm at the load location and the deflection curve with minus 0.827 times 10 to the power of minus 3 meters next to the load location.
Figure 6: Case V06 with L = 4 m and F = 1300 N downward at a = 1.2 m.

Analytical

Ra=F(L−a)L=910 NR_\mathrm{a}=\frac{F(L-a)}{L}=910\,\mathrm{N}

Rb=FaL=390 NR_\mathrm{b}=\frac{Fa}{L}=390\,\mathrm{N}

Mmax=Fa(L−a)L=1092 NmM_\mathrm{max}=\frac{Fa(L-a)}{L}=1092\,\mathrm{Nm}

wmin=−Fa (L2−a2)3/293 L EI=−0.827359 mmw_\mathrm{min}=-\frac{Fa\,(L^2-a^2)^{3/2}}{9\sqrt{3}\,L\,EI}=-0.827359\,\mathrm{mm}

at x=L−L2−a23=1.79697 mx=L-\sqrt{\frac{L^2-a^2}{3}}=1.79697\,\mathrm{m} for a<L/2a<L/2

SpanDraft

Ra=910 NR_\mathrm{a}=910\,\mathrm{N}

Rb=390 NR_\mathrm{b}=390\,\mathrm{N}

Mmax=1092 NmM_\mathrm{max}=1092\,\mathrm{Nm}

wmin=−0.827359 mmw_\mathrm{min}=-0.827359\,\mathrm{mm}

at x=1.79697 mx=1.79697\,\mathrm{m}

Fields between the stations

Extreme values alone do not show whether the field between the stations is correct. V06 has only three event stations: the two supports and the load location. The values in between come from the analytical curves of the pieces. Four of the evaluation points in the comparison:

xx in mww in mm, analyticalww in mm, SpanDraftMM in Nm, analyticalMM in Nm, SpanDraft
0.548−0.388837−0.388837498.68498.68
1.332−0.768186−0.7681861040.521040.52
2.472−0.722743−0.722743595.92595.92
3.492−0.281101−0.281101198.12198.12

Other orders of magnitude

The three variants of V04 check the same system at very different lengths, loads and stiffnesses. They are complete cases with both programs and all physical checks. The formulas from V04 apply unchanged. V04-S01 and V04-S03 show the same digits of the deflection because qL4/(EI)qL^4/(EI) grows by exactly the factor 10510^5 between them.

CaseLL in mqq in N/mEIEI in Nm²MmaxM_\mathrm{max} in Nm, analyticalMmaxM_\mathrm{max} in Nm, SpanDraftwminw_\mathrm{min} in m, analyticalwminw_\mathrm{min} in m, SpanDraft
V04-S010.161680.00750.0075−4.65030⋅10−8-4.65030\cdot10^{-8}−4.65030⋅10−8-4.65030\cdot10^{-8}
V04-S0216056007.57.5−1.39509⋅10−4-1.39509\cdot10^{-4}−1.39509⋅10−4-1.39509\cdot10^{-4}
V04-S03106001.68⋅1071.68\cdot10^775007500−4.65030⋅10−3-4.65030\cdot10^{-3}−4.65030⋅10−3-4.65030\cdot10^{-3}

How close the values are

The comparisons are rounded to six digits and therefore show equal digits. The actual check works with the unrounded values and the tolerance condition from the short version. The measure is the utilization aa of this tolerance. For a=1a=1, the tolerance would be exactly exhausted.

a=∣s−r∣absTol+relTol ∣r∣a=\frac{|s-r|}{\mathrm{absTol}+\mathrm{relTol}\,|r|}
(2)

The table gives the largest utilization over all compared values for each case. Extreme locations are not included, because there it is checked whether the SpanDraft location belongs to the set of reference locations.

CaseIndividual comparisonsLargest utilizationStatus
V01871.1⋅10−91.1\cdot10^{-9}passed
V02633.2⋅10−103.2\cdot10^{-10}passed
V03634.9⋅10−94.9\cdot10^{-9}passed
V04631.2⋅10−91.2\cdot10^{-9}passed
V05878.1⋅10−108.1\cdot10^{-10}passed
V06875.4⋅10−95.4\cdot10^{-9}passed
V04-S01632.7⋅10−112.7\cdot10^{-11}passed
V04-S02636.4⋅10−106.4\cdot10^{-10}passed
V04-S03631.7⋅10−81.7\cdot10^{-8}passed
Total6391.7⋅10−81.7\cdot10^{-8}passed

The utilization is calculated with the current solver against the same stored references as in the acceptance. For the other nine cases, no analytical reference is stored, although they too are in principle analytically solvable. They are checked by the chapter External reference solvers.

Sources

  1. Felix Udoeyo. Deflection of Beams: Geometric Methods. Structural Analysis, Section 7.3 Deflection by Method of Double Integration, Temple University Press / North Broad Press. 2020. Accessed .