Printer Friendly

Parallel-to-grain end-load capacity of round mortises in round and rectangular timbers.


Tests were conducted to determine end-wall load capacity of simulated tie beam mortises loaded parallel to the grain in double shear and their modes of failure. Test results indicate that load capacity of mortise end-walls in tie beams is related both to wood strength perpendicular to the grain and to wood shear strength parallel to the grain. Results also indicate that cleavage forces perpendicular to the grain in the walls of large mortises are inherently greater than in smaller mortises. Cross bolting the wall of the mortise substantially increases the load-carrying capacity of the mortise and decreases variability. Cross bolts should be located near the end of the tie beam. Differences in the strength of wood perpendicular to the grain also should be considered in selecting tie beams.

In timber frames constructed with round mortise and tenon joints, the load-carrying capacities of the tie beam to corner post joints (Fig. 1) are especially critical because a large proportion of the roof loads are transferred to these joints by end rafters and wall plates. Equally important are the load capacities of splice plate connections in the tie beam itself, Figure 2.


The load-carrying capacities of these connections are unknown, but they are expected to be limited by their location near the end of the tie beam. Related work with through bolt and dowel nut joints (Eckelman 2004) indicates that joint capacity can be increased either by locating the mortise farther from the end of the beam or by reinforcing the end-wall of the mortise with a cross bolt. Cross bolting is also effective (USDA 1999) for reinforcement of bolted end connector joints. Potentially, therefore, these joints could be designed to provide the resistance needed for highly loaded tie beams. Because the performance of the roof system depends upon the structural integrity of the tie beam joints, it is necessary to have rational estimates of their load capacities, information concerning their modes of failure, and knowledge of the factors that affect their behavior.

In a typical frame construction, the corner post tenon shown in Figure 1 is subjected to horizontal side thrust forces imposed by the wall plate and by the end rafter. The mortise wall of the tie beam, therefore, is subjected to single-shear forces. In contrast, in the tie beam splice joint shown in Figure 2, the mortise wall of the tie beam is subjected to double-shear forces. Therefore, the load capacity of the joint for both types of loading is needed. As a first step, however, only the mortise-wall load capacities of tie beam joint specimens loaded in double shear were investigated. The simpler double shear method of loading reduces variability and allows for better evaluation of the effectiveness of reinforcement methods. Thus, double shear end-connector tests were conducted. Specimens were used in which the tenon connector consisted of a single steel dowel of the same diameter as that of potential corner post tenons (Fig. 3). These specimens differed from conventional end-connector joints or tension splices in that the dowels were much larger in diameter and only one fastener was used.


These large diameter steel dowels provide a uniform bearing pressure on the end-wall of the mortise (Fig. 4) but produce larger splitting forces (Karlsen 1967). In addition, end distances were less than the 5 to 7 times bolt diameter specified for bolt connectors (USDA 1999).


Purpose and objectives

The purpose of the study was to determine the end-wall load capacities, behavior, and modes of failure of simulated tie beam mortises loaded parallel to the grain in double shear. Specific objectives were to determine the correlation between load-carrying capacities of the end-wall and tenon diameter; mortise end-distance; tie beam cross section; and wood species and to evaluate the effectiveness of cross bolting the mortise end-wall.

Specimen construction

Typical geometry of the specimen is shown in Figure 3. Wood species included eastern softwoods (Pinus, sp.), red pine (Pinus resinosa), southern yellow pine (Pinus, sp.), yellow-poplar (Liriodendron tulipifera), Douglas-fir (Pseudotsuga menziessii), red elm (Ulmus rubra), red oak (Quercus rubra), and white ash (Fraxinus alba). Specimens constructed of eastern softwoods (EP), preservative treated red pine (RP), no. 1 and no. 2 southern yellow pine (SYP), and Douglas-fir (D-fir) were 3.5 by 3.5 inches in cross section, whereas comparable yellow-poplar (Y-pop) specimens were 4 by 4 inches. These specimens had 2-inch diameter mortises and end-distances of 1.75, 2, 4, 5, or 6 inches where "end-distance" is the distance measured from the center of the mortise to the end of the specimen (Fig. 3). Specimens constructed of southern yellow pine were 5.5 by 5.5 inches, whereas yellow-poplar and red oak specimens were 6 by 6 inches. Such specimens had 3-inch mortises and end-distances of either 3 or 6 inches. Specimens constructed of yellow-poplar were 8 by 8 inches with 4-inch diameter mortises and 4- or 8-inch end-distances. The "load-end" of each specimen was constructed with a 2-inch mortise and was reinforced with either one or two cross bolts. Specimens were prepared in pairs--the test ends of half of the specimens were reinforced with one or two 0.375-inch diameter cross bolts (Fig. 2, Tables 1 to 3), whereas the test ends of the remaining half were not.

In addition, specimens with round cross sections were constructed from small-diameter red elm (R-Elm), white ash (W-Ash), and red oak (R-Oak) tree stems. These specimens had 2-inch diameter mortises and 4- and 6-inch mortise end-distances. Diameter of the specimens averaged 5 inches. Moisture content (MC) of the specimens varied from about 10 to 15 percent at time of test (Table 3). The load ends of all the round specimens were reinforced with one 0.375-inch diameter cross bolt; half of the test ends were also reinforced with one 0.375-inch diameter cross bolt, whereas the remaining half were not.

Except for the round stock, all specimens were cut from remnants of material or from a common stock of material that had been used in actual construction of three light building frames. All specimens contained juvenile wood. In fact, many were cut from timbers that had been sawn to include the heart of the tree. Except where noted, six specimens of each configuration were constructed. The material was initially stored outdoors in a covered shed; then it was brought indoors and conditioned to 8 percent MC before machining. The round material was also stored outdoors in a covered shed and subsequently brought indoors and allowed partial seasoning. Mortises were bored perpendicular to the plane of the major splits (if any) that had developed in the ends of the members. MCs shown in Table 3 were determined at time of test.

Test method

Specimens were attached to testing apparatus as illustrated in Figure 3. Load was applied to the "load-end" of the specimen through a 2-inch diameter steel dowel located 6 inches from the end of the specimen and to the "test-end" through 2-, 3-, or 4-inch diameter steel dowels at end distances of 1.75, 2, 3, 4, 5, or 6 inches. Testing was conducted in a universal testing machine at a rate of 0.05 inch/min. Ultimate load is reported as the load obtained at the time of catastrophic failure of the specimen--usually shear failure of the end-wall adjacent to the mortise boundary shear planes.


Results of the tests are given in Tables 1 through 3. To aid interpretation, results are also presented graphically in Figure 5.


The most common mode of failure of the mortise end-walls without cross bolts was cleavage of the end-wall in tension perpendicular to the grain with formation of an end split (Fig. 4), which led to partial relaxation of load. As testing continued, loading increased until shear failures parallel to the mortise boundary planes occurred (Fig. 4), either along one plane and then the other, or, simultaneously along both planes. In those cases in which failure occurred along only one shear plane, partial load relaxation again occurred, but as testing continued, loading increased until failure occurred along the other shear plane. Another mode of failure was tension of the mortise side walls in the softwood specimens; such failures, however, occurred only at high load levels and were very few.

The two modes of failure observed indicate that both the strength of the wood perpendicular to the grain and its shear strength parallel to the grain affect the load-carrying capacity of the joints. In most cases, cleavage of the end-wall occurred first, indicating that initial joint failure was most closely associated with tension perpendicular to the grain. It seems reasonable to assume that maximum cleavage forces occurred near the projected end-wall of the specimen. The interrelated deformations and bearing and bending forces that limit the ultimate load capacity of doweled joints are complex, but when end distances exceed critical values, it is postulated that initial cleavage would occur within the end-wall rather than at its projected end. Overall, it follows that maximum joint load capacity is obtained when cleavage capacity and shear capacity are equalized. Also, higher joint load capacity would be expected with wood species that have higher mechanical strength perpendicular to the grain. Joints constructed from yellow-poplar, therefore, would be expected to have higher load capacity than those constructed from northern pines, and those constructed from red elm would have higher capacity than those constructed from yellow-poplar. All of the specimens tested contained juvenile wood, however, which may have overshadowed species difference, particularly in the case of the softwoods. Variations in mechanical properties of juvenile wood among species, therefore, should be considered in the structural uses of small-diameter tree stems. The strengths of juvenile softwoods, for example, are substantially less than those of mature wood (Senti et al. 1986, Larson et al. 2001); however, reductions in strength may be less marked in hardwoods (Panshin and de Zeeuw 1970).

The effect of mortise end-distance on load-carrying capacity can be seen in Figure 5. Average ultimate load-carrying capacity of the 2-inch mortises in 3.5-inch members with 1.75-inch end-distances ranged from a low of 2,120 pounds for red pine to a high of 4,000 pounds for yellow-poplar. These joints correspond to construction in which the end of the tie beam would be flush with the outside edge of the corner post. The load-carrying capacity of such joints is limited by the short shear plane length of the mortise boundary wall, the thickness of the wall itself, and the depth of penetration of end splits into the mortise end-wall. The test results indicate that such joints might be satisfactory for small structures with low roof loads, but most likely they are not suitable for larger structures. Substantial increase in ultimate load-carrying capacity was observed when the mortise end-distance was increased to 4 inches. The load-carrying capacity of Douglas-fir, which had multiple small end checks, essentially doubled, e.g., from 2,552 pounds to 4,967 pounds, when the end-distance was increased from 1.75 to 4 inches. Similarly, the load capacity of the eastern softwoods, which did not contain end checks, increased from 2,150 pounds to 12,225 pounds. The ultimate load-carrying capacity of the southern yellow pine increased substantially, from 3,110 to 19,483 pounds; presumably, part of this increase was due to the fact that the specimens with 4-inch end-distance were constructed from no. 1 southern yellow pine, whereas the specimens with 1.75-inch end-distances were constructed from no. 2 southern yellow pine. A similar trend, with exceptions (presumably due to variations in material quality), was also observed as the end-distances of the yellow-poplar specimens with 2-inch mortises were increased from 1.75 to 6 inches--from 4,000 pounds for 1.75-inch end-distance to 20,213 pounds for 6-inch end-distance. These increases in load capacity result from better equalization of shear load and cleavage load capacities in the end-wall.

Similarly, increases in strength were observed for 3- and 4-inch mortises as end-distances increased. Specifically, the load capacity of yellow-poplar specimens with 3-inch mortises increased from 13,975 pounds to 18,250 pounds as the end-distance increased from 3 to 6 inches; likewise, the load capacity of comparable southern yellow pine specimens increased from 8,133 to 20,000 pounds. A small decrease in load capacity was observed with increases in end-distance from 4 to 6 inches in yellow-poplar specimens with 4-inch mortises. Presumably, the decrease in load capacity in these large specimens was due to substantial seasoning splits.

The increase in load capacity from the use of cross bolts is also shown in Figure 5. In the case of 3.5-inch square members with 1.75-inch end-distances, an average increase of 61 percent in load capacity was observed. These results are important in that the cross-bolted joints would most likely provide adequate resistance for small-building flame constructions. In addition, cross bolts provide a high degree of reliability against the weakening effects of end splits. Furthermore, cross bolts are increasingly effective in reinforcing against cleavage forces as the diameter of the mortises increase. In the case of 2-inch mortises, for example, on average, cross bolts increased the load capacity of the joints by 28 percent, whereas the increase in load capacity of 3- and 4-inch mortises was 47 percent and 127 percent, respectively. Thus, reinforcing tie beam mortises with cross bolts may be the solution to provide needed load capacities for large-diameter tenons.

As previously discussed, cleavage of the end-wall presumably originates on the projected external end-wall surface. In the case of 4-inch square yellow-poplar specimens with 2-inch mortises, the average load capacity of specimens with the cross bolt located 1-inch from the end of the specimens was 54 percent and 39 percent greater than those with cross bolts located 4 and 2.5 inches from the end of the specimen, respectively. These results tend to indicate that maximum cleavage forces do, in fact, occur near the projected end wall, or, end of the specimens. and, in practice, cross bolts should be located near the ends of tie beams to obtain maximum strength.

Test results also indicate that smaller diameter dowels generate smaller perpendicular to the grain forces in the end-wall of the mortise than do large dowels. For example, 5.5-inch square southern yellow pine specimens with 3-inch dowels, 6-inch end spacing, and two cross bolts, experienced failure of the wall of the 3-inch mortise in every case rather than the walls of the 2-inch load mortises. Identical results were obtained in the tests of the 8-inch square yellow-poplar specimens with 4-inch dowels, 6-inch end-distance, and two cross bolts.

Test results indicate that high load capacity can be obtained with round specimens provided mortises are bored perpendicular to the plane of major splits. Use of green material must thus be delayed until seasoning splits have developed.

Cross bolts greatly improve performance. On average, use of a single 0.375-inch diameter cross bolt increased load capacity by 44 percent. The load capacity of red elm specimens compared to those of white ash specimens indicates that load capacity is closely related to tension perpendicular to the grain. A similar result was observed for red oak after differences in mortise depth were taken into account.


Test results indicate that load capacity of mortise end-walls in tie beams is related both to the wood strength perpendicular to the grain and to the wood shear strength parallel to the grain. Thus, a white ash tie beam would be expected to have greater load capacity than a yellow-poplar tie beam. Load capacity of the mortise joint is also closely related to the "end-distance" of the mortise. Thus, it is advantageous to allow the end of the tie beam to project as far as is feasible beyond the mortise. Tension forces perpendicular to the grain in the end-walls of large mortises are inherently greater than in smaller mortises. Thus, the load capacities of large members--with large mortises--may be less than intuitively expected.

Cross bolting the wall of the mortise substantially increases load-carrying capacity of the mortise. Bolts should be located near the end of the tie beam. Highest load capacities were obtained with cross bolts located 1-inch from the end of the beam. Cross bolts are particularly effective in reinforcing large mortises; they also decrease variability and increase reliability in beams with smaller mortises.

Results indicate that mortise end-connector joints--even with end-wall distances no greater than the width of the member--have sufficient load-carrying capacity for use in many tie beam joint applications, provided the mortise wall is reinforced with cross bolts. Differences in the strength of wood perpendicular to the grain should be considered in selecting tie beams.

Literature cited

Eckelman, C. 2004. Exploratory study of high-strength low-cost through-bolt with cross pipe and nut construction for roundwood and squared timber frame construction. Forest Prod. J. 54(12):29-37.

Karlsen, G.G. 1967. Wooden Structures. MIR Publishers. Moscow. 638 pp.

Larson, P.R., D.E. Kretschmann, A. Clark, III, and J.G. Isebrands. 2001. Formation and properties of juvenile wood in southern pines. Gen. Tech. Rept. FPL-GTR-129. USDA Forest Serv., Forest Prod. Lab., Madison, Wisconsin.

Panshin, A.J. and C. de Zeeuw. 1970. Textbook of Wood Tech., Vol. I. MacGraw-Hill. 705 pp.

Senft, J.F., M.J. Quanci, and B.A. Bendtsen. 1986. Property profile of 60-year old Douglas-fir. In: Juvenile Wood--What Does It Mean to Forest Management and Forest Products? Forest Products Soc. Proc. 47309.

USDA Forest Service, Forest Products Lab, 1999. Wood Handbook: Wood as an Engineering Material. Forest Prod. Soc. Madison, Wisconsin.

C.A. Eckelman * E. Haviarova * H. Akcay

The authors are, respectively, Professor, Assistant Professor, and Graduate Student, Purdue Univ., West Lafayette, Indiana (,, This paper was received for publication in April 2006. Article No. 10187.

* Forest Products Society Member.
Table 1.--Test results for 2-inch mortises in rectangular
specimens at 8 percent MC.


 Test No of Wood Depth Mort End
set no. spec. species by width diam. dist.


2-1 3 RP 3.5 2 1.75
2-2 3 RP 3.5 2 1.75
2-3 3 EP 3.5 2 1.75
2-4 3 EP 3.5 2 1.75
2-5 3 EP 3.5 2 4
2-6 3 EP 3.5 2 4
2-7 3 D-fir 3.5 2 1.75
2-8 3 D-fir 3.5 2 1.75
2-9 3 D-fir 3.5 2 4
2-10 3 D-fir 3.5 2 4
2-11 3 SYP 3.5 2 1.75
2-12 3 SYP 3.5 2 1.75
2-13 3 SYP 3.5 2 4
2-14 4 SYP 3.5 2 4
2-15 3 Y-pop 4 2 1.75
2-16 3 Y-pop 4 2 1.75
2-17 6 Y-pop 4 2 2
2-18 6 Y-pop 4 2 2
2-19 3 Y-pop 4 2 3
2-20 2 Y-pop 4 2 3
2-21 1 Y-pop 4 2 4
2-22 4 Y-pop 4 2 4
2-23 4 Y-pop 4 2 5
2-26 3 Y-pop 4 2 5
2-25 4 Y-pop 4 2 6
2-26 3 Y-pop 4 2 6
2-27 3 Y-pop 4 2 6
2-28 3 Y-pop 4 2 6
2-29 3 Y-pop 4 2 6

 Cross Bolts

 Test No. of End
set no. diam. dist. Avg. SD

 (in) (lb)

2-1 none na 2,120 308
2-2 1 by 0.25 0.75 3,193 275
2-3 none na 2,150 330
2-4 1 by 0.25 0.75 3,632 558
2-5 none na 12,225 2,840
2-6 1 by 0.25 2 10,025 2,762
2-7 none na 2,552 305
2-8 2 by 0.25 0.75 3,737 748
2-9 none na 4,967 1,020
2-10 1 by 0.25 2 8,100 889
2-11 none na 3,110 108
2-12 2 by 0.25 5,087 298
2-13 none na 19,483 5,290
2-14 1 by 0.25 1.50 13,662 2,853
2-15 none na 4,000 691
2-16 1 by 0.25 0.75 7,075 229
2-17 none na 7,025 1,346
2-18 1 by 0.375 0.75 8,928 1,632
2-19 na na 13,317 2,760
2-20 1 by 0.375 1 18,650 4,950
2-21 none na 11,000
2-22 0.375 1 14,062 1,098
2-23 none na 24,113 8,259
2-26 1 by 0.375 1 33,150 2,087
2-25 none na 20,213 9,451
2-26 1 by 0.375 1 31,967 2,926
2-27 1 by 0.375 1 35,817 6,264
2-28 1 by 0.375 2.50 25,717 4,283
2-29 1 by 0.375 4 23,283 3,514

Table 2.--Test results for 3- and 4-inch mortises at test end and
2-inch mortises at load end in square specimens at 8 percent MC
(red oak @ 59%).


Test No. of Wood Depth Mort. diam. End dist.
set no. spec. Spec. by width T-end T-end


 3-1 4 Y-pop 6 3 3
 3-2 3 Y-pop 6 3 3
 3-3 3 Y-pop 6 3 6
 3-4 3 Y-pop 6 3 6
 3-5 3 R-Oak 6 3 3
 3-6 3 R-Oak 6 3 3
 3-7 3 SYP 5.5 3 3
 3-8 2 SYP 5.5 3 3
 3-9 3 SYP 5.5 3 6
3-10 2 SYP 5.5 3 6
3-11 3 SYP 5.5 3 6
 4-1 3 Y-pop 8 4 4
 4-2 3 Y-pop 8 4 4
 4-3 3 Y-pop 8 4 6
 4-4 3 Y-pop 8 4 6

 Cross Bolts

Test No. by End
set no. diam. dist. Avg. SD

 (in) (lb)

 3-1 none na 13,975 1,765
 3-2 2 by 0.5 1 19,033 3,126
 3-3 none na 18,250 5,360
 3-4 1 by 0.5 1 24,967 2,053
 3-5 none na 15,783 818
 3-6 2 by 0.375 1.5 25,567 2,850
 3-7 na 8,133 723
 3-8 1 by 0.375 1 13,600 636
 3-9 none na 20,000 2,081
3-10 2 by 0.375 1.5 28,150 3,161
3-11 2 by 0.375 1.5 30,000 na
 4-1 none na 18,383 3,393
 4-2 2 by 0.375 1 29,300 1,825
 4-3 none na 15,400 568
 4-4 2 by 0.5 1 47,433 7,795

Table 3.--Round specimens with 2-inch mortises.

Test No. of Wood MC Diam. Along Diam across
set no. spec. species tenon (avg/SD) tenon (avg/SD)

 (%) (in)

R-1 8 W-Ash 15.5 4.570 4.788
 0.506 0.617
R-2 9 W-Ash 15.4 4.708 4.830
 0.4 0.406 0.436
R-3 3 R-Elm 14.6 4.5 4.960
 0.8 0.573 0.408
R-4 4 R-Elm 10.0 4.125 4.600
 0.375 0.376
R-5 4 R-Elm 9.8 4.450 4.657
 0.4 0.676 0.592
R-6 3 R-Oak 14.0 3.715 4.468
 0.4 0.261 0.269
R-7 8 R-Oak 12.7 4.174 4.286
 1.8 0.319 0.292

 Cross Bolts

Test Mort. Shear No. by End Ult. Load
set no. diam. plane length diam. dist. (avg/SD)

 (in) (lbs)

R-1 2 6 none na 19,926
R-2 2 6 1 by 0.375 1 29,800
R-3 2 4 none na 13,550
R-4 2 4 1 by 0.375 2 18,950
R-5 2 6 none na 27,462
R-6 2 6 none na 14,533
R-7 2 6 1 by 0.375 2 20,894
COPYRIGHT 2007 Forest Products Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2007 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Technical Note
Author:Eckelman, C.A.; Haviarova, E.; Akcay, H.
Publication:Forest Products Journal
Geographic Code:1USA
Date:Apr 1, 2007
Previous Article:Modeling lumber value recovery in relation to selected tree characteristics in black spruce using the Optitek sawing simulator.
Next Article:Marine borer degradation of treated and untreated parallel strand lumber panels.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters